A practical introduction to Random Forest for genetic association studies in ecology and evolution

Consequences of domestication in eastern oyster: Insights from whole genomic analyses

Selective breeding for production traits has yielded relatively rapid successes with high-fecundity aquaculture species. Discovering the genetic changes associated with selection is an important goal for understanding adaptation and can also facilitate better predictions about the likely fitness of selected strains if they escape aquaculture farms. Here, we hypothesize domestication as a genetic change induced by inadvertent selection in culture. Our premise is that standardized culture protocols generate parallel domestication effects across independent strains. Using eastern oyster as a model and a newly developed 600K SNP array, this study tested for parallel domestication effects in multiple independent selection lines compared with their progenitor wild populations. A single contrast was made between pooled selected strains (1-17 generations in culture) and all wild progenitor samples combined. Population structure analysis indicated rank order levels of differentiation as [wild - wild] < [wild - cultured] < [cultured - cultured]. A genome scan for parallel adaptation to the captive environment applied two methodologically distinct outlier tests to the wild versus selected strain contrast and identified a total of 1174 candidate SNPs. Contrasting wild versus selected strains revealed the early evolutionary consequences of domestication in terms of genomic differentiation, standing genetic diversity, effective population size, relatedness, runs of homozygosity profiles, and genome-wide linkage disequilibrium patterns. Random Forest was used to identify 37 outlier SNPs that had the greatest discriminatory power between bulked wild and selected oysters. The outlier SNPs were in genes enriched for cytoskeletal functions, hinting at possible traits under inadvertent selection during larval culture or pediveliger setting at high density. This study documents rapid genomic changes stemming from hatchery-based cultivation of eastern oysters, identifies candidate loci responding to domestication in parallel among independent aquaculture strains, and provides potentially useful genomic resources for monitoring interbreeding between farm and wild oysters.

Tropical peanut maturation scale for harvesting seeds with superior quality

Determining the moment for harvesting the tropical peanut with a focus on superior seed quality is not an easy task. Particularities such as indeterminate flowering, underground fruiting and uneven maturation further increase this technical challenge. It is in this context that we aim to investigate harvest indicators based on the maturation and late maturation phases of tropical peanuts to obtain seeds with superior physiological and health quality. The plants were grown in field conditions and their development stages were carefully monitored until seed production. The water content, dry weight, germination capacity, desiccation tolerance, vigor, longevity, and seed pathogens were evaluated throughout these stages. We showed that seeds from early stages (R5 and R6) did not fully tolerate desiccation and were highly sensitive to pathogen contamination after storage (Aspergillus, Penicillium, and Bacteria). At late stages (R7, R8, and R9), the seeds had optimized vigor, longevity and bioprotection against fungi and thermal stress. The peanut maturation scale for tropical agriculture provides unique harvesting guidelines that make it possible to monitor the plants' development stages with a focus on producing superior quality seeds.

Variable parallelism in the genomic basis of age at maturity across spatial scales in Atlantic Salmon

Complex traits often exhibit complex underlying genetic architectures resulting from a combination of evolution from standing variation, hard and soft sweeps, and alleles of varying effect size. Increasingly, studies implicate both large-effect loci and polygenic patterns underpinning adaptation, but the extent that common genetic architectures are utilized during repeated adaptation is not well understood. Sea age or age at maturation represents a significant life history trait in Atlantic Salmon (Salmo salar), the genetic basis of which has been studied extensively in European Atlantic populations, with repeated identification of large-effect loci. However, the genetic basis of sea age within North American Atlantic Salmon populations remains unclear, as does the potential for a parallel trans-Atlantic genomic basis to sea age. Here, we used a large single-nucleotide polymorphism (SNP) array and low-coverage whole-genome resequencing to explore the genomic basis of sea age variation in North American Atlantic Salmon. We found significant associations at the gene and SNP level with a large-effect locus (vgll3) previously identified in European populations, indicating genetic parallelism, but found that this pattern varied based on both sex and geographic region. We also identified nonrepeated sets of highly predictive loci associated with sea age among populations and sexes within North America, indicating polygenicity and low rates of genomic parallelism. Despite low genome-wide parallelism, we uncovered a set of conserved molecular pathways associated with sea age that were consistently enriched among comparisons, including calcium signaling, MapK signaling, focal adhesion, and phosphatidylinositol signaling. Together, our results indicate parallelism of the molecular basis of sea age in North American Atlantic Salmon across large-effect genes and molecular pathways despite population-specific patterns of polygenicity. These findings reveal roles for both contingency and repeated adaptation at the molecular level in the evolution of life history variation.

Temperature and local anthropogenic pressures limit stony coral assemblage viability in southeast Florida

Climate change is viewed as the primary threat to coral reefs, with local pressures exacerbating coral cover decline. The consensus is that improving water quality may increase resilience, but disentangling water quality and temperature impacts is difficult. We used distance-based linear models and random forests to analyze spatiotemporal variation in benthic community structure and interannual changes in the coral assemblage, in relation to specific environmental metrics in Southeast Florida. Temperature accounted for most of the variation, recruitment doubled and interannual increases in coral abundance tripled when mean annual temperature reached 27 °C, until maximum temperatures exceeded 31 °C. Benefits associated with warmer temperatures were negated by poor water quality, as nutrient enrichment was related to increased macroalgal cover, reduced coral recruitment and higher coral partial mortality. We suggest reducing local pressures will contribute to reduced macroalgae and enhance coral recovery, but that temperature is the predominant influence on coral assemblages.

Maximising environmental pressure-response relationship signals from diatom-based metabarcoding in rivers

DNA metabarcoding has been performed on a large number of river phytobenthos samples collected from the UK, using rbcL primers optimised for diatoms. Within this dataset the composition of non-diatom sequence reads was studied and the effect of including these in models for evaluating the nutrient gradient was assessed. Whilst many non-diatom taxonomic groups were detected, few contained the full diversity expected in riverine environments. This may be due to the performance of the current primers in characterising the wider phytobenthic community and influenced by the sampling method employed, as both were developed specifically for diatoms. Nevertheless, the study identified considerable diversity in some groups, e.g. Eustigmatophyceae and a wider distribution than previously thought for freshwater Phaeophyceae. These results offer a strong case for the benefits of metabarcoding for expanding knowledge of aquatic biodiversity in the UK and elsewhere. Many of the ASVs associated with non-diatoms showed significant pressure responses; however, models that included non-diatoms had similar predictive strength to those based on diatoms alone. Whilst limitations of the primers for assessing non-diatoms may play a role in explaining these results, the diatoms provide a strong signal along the nutrient gradient and other algae, therefore, add little unique information. We recommend that future developments should use ASVs to calculate metrics, with links to reference databases made as a final step to generate lists of taxa to support interpretation. Any further exploration of the potential of non-diatoms would benefit from access to a well-curated reference database, similar to diat.barcode. Such a database does not yet exist, and we caution against the indiscriminate use of NCBI GenBank as a taxonomic resource as many rbcL sequences deposited have not been curated.

Source identification of sedimentary organic carbon in coastal wetlands of the western Bohai Sea

Coastal wetlands play a vital role in mitigating climate change, yet the characteristics of buried organic carbon (OC) and carbon cycling are limited due to difficulties in assessing the composition of OC from different sources (allochthonous vs. autochthonous). In this study, we analyzed the total organic carbon (TOC) to total nitrogen (TN) ratio (C/N), stable carbon isotope (δ13C) composition, and n-alkane content to distinguish different sources of OC in the surface sediments of the coastal wetlands on the western coast of the Bohai Sea. The coupling of the C/N ratio with δ13C and n-alkane biomarkers has been proved to be an effective tool for revealing OC sources. The three end-member Bayesian mixing model based on coupling C/N ratios with δ13C showed that the sedimentary OC was dominated by the contribution of terrestrial particulate organic matter (POM), followed by freshwater algae and marine phytoplankton, with relative contributions of 47 ± 21 %, 41 ± 18 % and 12 ± 17 %, respectively. The relative contributions of terrestrial plants, aquatic macrophytes and marine phytoplankton assessed by n-alkanes were 56 ± 8 %, 35 ± 9 % and 9 ± 5 % in the study area, respectively. The relatively high salinity levels and strong hydrodynamic conditions of the Beidagang Reservoir led to higher terrestrial plants source and lower aquatic macrophytes source than these of Qilihai Reservoir based on the assessment of n-alkanes. Both methods showed that sedimentary OC was mainly derived from terrestrial sources (plant-dominated), suggesting that vegetation plays a crucial role in storing carbon in coastal wetlands, thus, the coastal vegetation management needs to be strengthened in the future. Our findings provide insights into the origins and dynamics of OC in coastal wetlands on the western coast of the Bohai Sea and a significant scientific basis for future monitoring of the blue carbon budget balance in coastal wetlands.

A machine learning approach for modeling the occurrence of the major intermediate hosts for schistosomiasis in East Africa

Schistosomiasis, a prevalent water-borne disease second only to malaria, significantly impacts impoverished rural communities, primarily in Sub-Saharan Africa where over 90% of the severely affected population resides. The disease, majorly caused by Schistosoma mansoni and S. haematobium parasites, relies on freshwater snails, specifically Biomphalaria and Bulinus species, as crucial intermediate host (IH) snails. Targeted snail control is advisable, however, there is still limited knowledge about the community structure of the two genera especially in East Africa. Utilizing a machine learning approach, we employed random forest to identify key features influencing the distribution of both IH snails in this region. Our results reveal geography and climate as primary factors for Biomphalaria, while Bulinus occurrence is additionally influenced by soil clay content and nitrogen concentration. Favorable climate conditions indicate a high prevalence of IHs in East Africa, while the intricate connection with geography might signify either dispersal limitations or environmental filtering. Predicted probabilities demonstrate non-linear patterns, with Bulinus being more likely to occur than Biomphalaria in the region. This study provides foundational framework insights for targeted schistosomiasis prevention and control strategies in the region, assisting health workers and policymakers in their efforts.

Gut-resident microorganisms and their genes are associated with cognition and neuroanatomy in children

Emerging evidence implicates gut microbial metabolism in neurodevelopmental disorders, but its influence on typical neurodevelopment has not been explored in detail. We investigated the relationship between the microbiome and neuroanatomy and cognition of 381 healthy children, demonstrating that differences in microbial taxa and genes are associated with overall cognitive function and the size of brain regions. Using a combination of statistical and machine learning models, we showed that species including Alistipes obesi, Blautia wexlerae, and Ruminococcus gnavus were enriched or depleted in children with higher cognitive function scores. Microbial metabolism of short-chain fatty acids was also associated with cognitive function. In addition, machine models were able to predict the volume of brain regions from microbial profiles, and taxa that were important in predicting cognitive function were also important for predicting individual brain regions and specific subscales of cognitive function. These findings provide potential biomarkers of neurocognitive development and may enable development of targets for early detection and intervention.

Genomic Divergence and the Evolution of Ecotypes in Bottlenose Dolphins (Genus Tursiops)

Climatic changes have caused major environmental restructuring throughout the world's oceans. Marine organisms have responded to novel conditions through various biological systems, including genomic adaptation. Growing accessibility of next-generation DNA sequencing methods to study nonmodel species has recently allowed genomic changes underlying environmental adaptations to be investigated. This study used double-digest restriction-site associated DNA (ddRAD) sequence data to investigate the genomic basis of ecotype formation across currently recognized species and subspecies of bottlenose dolphins (genus Tursiops) in the Southern Hemisphere. Subspecies-level genomic divergence was confirmed between the offshore common bottlenose dolphin (T. truncatus truncatus) and the inshore Lahille's bottlenose dolphin (T. t. gephyreus) from the southwestern Atlantic Ocean (SWAO). Similarly, subspecies-level divergence is suggested between inshore (eastern Australia) Indo-Pacific bottlenose dolphin (T. aduncus) and the proposed Burrunan dolphin (T. australis) from southern Australia. Inshore bottlenose dolphin lineages generally had lower genomic diversity than offshore lineages, a pattern particularly evident for T. t. gephyreus, which showed exceptionally low diversity. Genomic regions associated with cardiovascular, musculoskeletal, and energy production systems appear to have undergone repeated adaptive evolution in inshore lineages across the Southern Hemisphere. We hypothesize that comparable selective pressures in the inshore environment drove similar adaptive responses in each lineage, supporting parallel evolution of inshore bottlenose dolphins. With climate change altering marine ecosystems worldwide, it is crucial to gain an understanding of the adaptive capacity of local species and populations. Our study provides insights into key adaptive pathways that may be important for the long-term survival of cetaceans and other organisms in a changing marine environment.

Spatial distribution and temporal dynamics of invasive and native mosquitoes in a large Mediterranean city

Mosquitoes, including invasive species like the Asian tiger mosquito Aedes albopictus, alongside native species Culex pipiens s.l., pose a significant nuisance to humans and serve as vectors for mosquito-borne diseases in urban areas. Understanding the impact of water infrastructure characteristics, climatic conditions, and management strategies on mosquito occurrence and effectiveness of control measures to assess their implications on mosquito occurrence is crucial for effective vector control. In this study, we examined data collected during the local vector control program in Barcelona, Spain, focusing on 234,225 visits to 31,334 different sewers, as well as 1817 visits to 152 fountains between 2015 and 2019. We investigated both the colonization and recolonization processes of mosquito larvae within these water infrastructures. Our findings revealed higher larval presence in sandbox-sewers compared to siphonic or direct sewers, and the presence of vegetation and the use of naturalized water positively influenced larval occurrence in fountains. The application of larvicidal treatment significantly reduced larvae presence; however, recolonization rates were negatively affected by the time elapsed since treatment. Climatic conditions played a critical role in the colonization and recolonization of sewers and urban fountains, with mosquito occurrence exhibiting non-linear patterns and, generally, increasing at intermediate temperatures and accumulated rainfall levels. This study emphasizes the importance of considering sewers and fountains characteristics and climatic conditions when implementing vector control programs to optimize resources and effectively reduce mosquito populations.

Predicting Environmental and Ecological Drivers of Human Population Structure

Landscape, climate, and culture can all structure human populations, but few existing methods are designed to simultaneously disentangle among a large number of variables in explaining genetic patterns. We developed a machine learning method for identifying the variables which best explain migration rates, as measured by the coalescent-based program MAPS that uses shared identical by descent tracts to infer spatial migration across a region of interest. We applied our method to 30 human populations in eastern Africa with high-density single nucleotide polymorphism array data. The remarkable diversity of ethnicities, languages, and environments in this region offers a unique opportunity to explore the variables that shape migration and genetic structure. We explored more than 20 spatial variables relating to landscape, climate, and presence of tsetse flies. The full model explained ∼40% of the variance in migration rate over the past 56 generations. Precipitation, minimum temperature of the coldest month, and elevation were the variables with the highest impact. Among the three groups of tsetse flies, the most impactful was fusca which transmits livestock trypanosomiasis. We also tested for adaptation to high elevation among Ethiopian populations. We did not identify well-known genes related to high elevation, but we did find signatures of positive selection related to metabolism and disease. We conclude that the environment has influenced the migration and adaptation of human populations in eastern Africa; the remaining variance in structure is likely due in part to cultural or other factors not captured in our model.

Prediction of detailed blood metabolic profile using milk infrared spectra and machine learning methods in dairy cattle

The adoption of preventive management decisions is crucial to dealing with metabolic impairments in dairy cattle. Various serum metabolites are known to be useful indicators of the health status of cows. In this study, we used milk Fourier-transform mid-infrared (FTIR) spectra and various machine learning (ML) algorithms to develop prediction equations for a panel of 29 blood metabolites, including those related to energy metabolism, liver function/hepatic damage, oxidative stress, inflammation/innate immunity, and minerals. For most traits, the data set comprised observations from 1,204 Holstein-Friesian dairy cows belonging to 5 herds. An exception was represented by β-hydroxybutyrate prediction, which contained observations from 2,701 multibreed cows pertaining to 33 herds. The best predictive model was developed using an automatic ML algorithm that tested various methods, including elastic net, distributed random forest, gradient boosting machine, artificial neural network, and stacking ensemble. These ML predictions were compared with partial least squares regression, the most commonly used method for FTIR prediction of blood traits. Performance of each model was evaluated using 2 cross-validation (CV) scenarios: 5-fold random (CV_r) and herd-out (CV_h). We also tested the best model's ability to classify values precisely in the 2 extreme tails, namely, the 25th (Q25) and 75th (Q75) percentiles (true-positive prediction scenario). Compared with partial least squares regression, ML algorithms achieved more accurate performance. Specifically, elastic net increased the R² value from 5% to 75% for CV_r and 2% to 139% for CV_h, whereas the stacking ensemble increased the R² value from 4% to 70% for CV_r and 4% to 150% for CV_h. Considering the best model, with the CV_r scenario, good prediction accuracies were obtained for glucose (R² = 0.81), urea (R² = 0.73), albumin (R² = 0.75), total reactive oxygen metabolites (R² = 0.79), total thiol groups (R² = 0.76), ceruloplasmin (R² = 0.74), total proteins (R² = 0.81), globulins (R² = 0.87), and Na (R² = 0.72). Good prediction accuracy in classifying extreme values was achieved for glucose (Q25 = 70.8%, Q75 = 69.9%), albumin (Q25 = 72.3%), total reactive oxygen metabolites (Q25 = 75.1%, Q75 = 74%), thiol groups (Q75 = 70.4%), total proteins (Q25 = 72.4%, Q75 = 77.2.%), globulins (Q25 = 74.8%, Q75 = 81.5%), and haptoglobin (Q75 = 74.4%). In conclusion, our study shows that FTIR spectra can be used to predict blood metabolites with relatively good accuracy, depending on trait, and are a promising tool for large-scale monitoring.

Fire frequency, as well as stress response and developmental gene control serotiny level variation in a widespread pioneer Mediterranean conifer, Pinus halepensis

Many plants undergo adaptation to fire. Yet, as global change is increasing fire frequency worldwide, our understanding of the genetics of adaptation to fire is still limited. We studied the genetic basis of serotiny (the ability to disseminate seeds exclusively after fire) in the widespread, pioneer Mediterranean conifer Pinus halepensis Mill., by linking individual variation in serotiny presence and level to fire frequency and to genetic polymorphism in natural populations. After filtering steps, 885 single nucleotide polymorphisms (SNPs) out of 8000 SNPs used for genotyping were implemented to perform an in situ association study between genotypes and serotiny presence and level. To identify serotiny‐associated loci, we performed random forest analyses of the effect of SNPs on serotiny levels, while controlling for tree size, frequency of wildfires, and background environmental parameters. Serotiny showed a bimodal distribution, with serotinous trees more frequent in populations exposed to fire in their recent history. Twenty‐two SNPs found in genes involved in stress tolerance were associated with the presence‐absence of serotiny while 37 found in genes controlling for flowering were associated with continuous serotiny variation. This study shows the high potential of P. halepensis to adapt to changing fire regimes, benefiting from a large and flexible genetic basis of trait variation.

Neuron numbers link innovativeness with both absolute and relative brain size in birds

A longstanding issue in biology is whether the intelligence of animals can be predicted by absolute or relative brain size. However, progress has been hampered by an insufficient understanding of how neuron numbers shape internal brain organization and cognitive performance. On the basis of estimations of neuron numbers for 111 bird species, we show here that the number of neurons in the pallial telencephalon is positively associated with a major expression of intelligence: innovation propensity. The number of pallial neurons, in turn, is greater in brains that are larger in both absolute and relative terms and positively covaries with longer post-hatching development periods. Thus, our analyses show that neuron numbers link cognitive performance to both absolute and relative brain size through developmental adjustments. These findings help unify neuro-anatomical measures at multiple levels, reconciling contradictory views over the biological significance of brain expansion. The results also highlight the value of a life history perspective to advance our understanding of the evolutionary bases of the connections between brain and cognition.

Deciphering Pleiotropic Signatures of Regulatory SNPs in Zea mays L. Using Multi-Omics Data and Machine Learning Algorithms

Maize is one of the most widely grown cereals in the world. However, to address the challenges in maize breeding arising from climatic anomalies, there is a need for developing novel strategies to harness the power of multi-omics technologies. In this regard, pleiotropy is an important genetic phenomenon that can be utilized to simultaneously enhance multiple agronomic phenotypes in maize. In addition to pleiotropy, another aspect is the consideration of the regulatory SNPs (rSNPs) that are likely to have causal effects in phenotypic development. By incorporating both aspects in our study, we performed a systematic analysis based on multi-omics data to reveal the novel pleiotropic signatures of rSNPs in a global maize population. For this purpose, we first applied Random Forests and then Markov clustering algorithms to decipher the pleiotropic signatures of rSNPs, based on which hierarchical network models are constructed to elucidate the complex interplay among transcription factors, rSNPs, and phenotypes. The results obtained in our study could help to understand the genetic programs orchestrating multiple phenotypes and thus could provide novel breeding targets for the simultaneous improvement of several agronomic traits.

IIb-RAD-sequencing coupled with random forest classification indicates regional population structuring and sex-specific differentiation in salmon lice (Lepeophtheirus salmonis)

The aquaculture industry has been dealing with salmon lice problems forming serious threats to salmonid farming. Several treatment approaches have been used to control the parasite. Treatment effectiveness must be optimized, and the systematic genetic differences between subpopulations must be studied to monitor louse species and enhance targeted control measures. We have used IIb-RAD sequencing in tandem with a random forest classification algorithm to detect the regional genetic structure of the Norwegian salmon lice and identify important markers for sex differentiation of this species. We identified 19,428 single nucleotide polymorphisms (SNPs) from 95 individuals of salmon lice. These SNPs, however, were not able to distinguish the differential structure of lice populations. Using the random forest algorithm, we selected 91 SNPs important for geographical classification and 14 SNPs important for sex classification. The geographically important SNP data substantially improved the genetic understanding of the population structure and classified regional demographic clusters along the Norwegian coast. We also uncovered SNP markers that could help determine the sex of the salmon louse. A large portion of the SNPs identified to be under directional selection was also ranked highly important by random forest. According to our findings, there is a regional population structure of salmon lice associated with the geographical location along the Norwegian coastline.

Whole-Genome Resequencing to Evaluate Life History Variation in Anadromous Migration of Oncorhynchus mykiss

Anadromous fish experience physiological modifications necessary to migrate between vastly different freshwater and marine environments, but some species such as Oncorhynchus mykiss demonstrate variation in life history strategies with some individuals remaining exclusively resident in freshwater, whereas others undergo anadromous migration. Because there is limited understanding of genes involved in this life history variation across populations of this species, we evaluated the genomic difference between known anadromous (n = 39) and resident (n = 78) Oncorhynchus mykiss collected from the Klickitat River, WA, USA, with whole-genome resequencing methods. Sequencing of these collections yielded 5.64 million single-nucleotide polymorphisms that were tested for significant differences between resident and anadromous groups along with previously identified candidate gene regions. Although a few regions of the genome were marginally significant, there was one region on chromosome Omy12 that provided the most consistent signal of association with anadromy near two annotated genes in the reference assembly: COP9 signalosome complex subunit 6 (CSN6) and NACHT, LRR, and PYD domain–containing protein 3 (NLRP3). Previously identified candidate genes for anadromy within the inversion region of chromosome Omy05 in coastal steelhead and rainbow trout were not informative for this population as shown in previous studies. Results indicate that the significant region on chromosome Omy12 may represent a minor effect gene for male anadromy and suggests that this life history variation in Oncorhynchus mykiss is more strongly driven by other mechanisms related to environmental rearing such as epigenetic modification, gene expression, and phenotypic plasticity. Further studies into regulatory mechanisms of this trait are needed to understand drivers of anadromy in populations of this protected species.

Modeling solubility of CO2–N2 gas mixtures in aqueous electrolyte systems using artificial intelligence techniques and equations of state

Determining the solubility of non-hydrocarbon gases such as carbon dioxide (CO₂) and nitrogen (N₂) in water and brine is one of the most controversial challenges in the oil and chemical industries. Although many researches have been conducted on solubility of gases in brine and water, very few researches investigated the solubility of power plant flue gases (CO₂–N₂ mixtures) in aqueous solutions. In this study, using six intelligent models, including Random Forest, Decision Tree (DT), Gradient Boosting-Decision Tree (GB-DT), Adaptive Boosting-Decision Tree (AdaBoost-DT), Adaptive Boosting-Support Vector Regression (AdaBoost-SVR), and Gradient Boosting-Support Vector Regression (GB-SVR), the solubility of CO₂–N₂ mixtures in water and brine solutions was predicted, and the results were compared with four equations of state (EOSs), including Peng–Robinson (PR), Soave–Redlich–Kwong (SRK), Valderrama–Patel–Teja (VPT), and Perturbed-Chain Statistical Associating Fluid Theory (PC-SAFT). The results indicate that the Random Forest model with an average absolute percent relative error (AAPRE) value of 2.8% has the best predictions. The GB-SVR and DT models also have good precision with AAPRE values of 6.43% and 7.41%, respectively. For solubility of CO₂ present in gaseous mixtures in aqueous systems, the PC-SAFT model, and for solubility of N₂, the VPT EOS had the best results among the EOSs. Also, the sensitivity analysis of input parameters showed that increasing the mole percent of CO₂ in gaseous phase, temperature, pressure, and decreasing the ionic strength increase the solubility of CO₂–N₂ mixture in water and brine solutions. Another significant issue is that increasing the salinity of brine also has a subtractive effect on the solubility of CO₂–N₂ mixture. Finally, the Leverage method proved that the actual data are of excellent quality and the Random Forest approach is quite reliable for determining the solubility of the CO₂–N₂ gas mixtures in aqueous systems.

Riverscape Genomics Clarifies Neutral and Adaptive Evolution in an Amazonian Characin Fish (Triportheus albus)

Understanding the role of natural selection in the evolution of wild populations is challenging due to the spatial complexity of natural systems. The richest diversity of freshwater fishes in the world is found in the Amazon Basin, a system where marked hydrochemical differences exist at the interface of major rivers with distinct “water colors” (i.e., black, white, and clear water). We hypothesize that divergent natural selection associated with these “aquatic ecotones” influences population-level adaptive divergence in the non-migratory Amazonian fish fauna. This hypothesis was tested using a landscape genomics framework to compare the relative contribution of environmental and spatial factors to the evolutionary divergence of the Amazonian characin fish Triportheus albus. The framework was based on spatial data, in situ hydrochemical measurements, and 15,251 filtered SNPs (single nucleotide polymorphisms) for T. albus sampled from three major Amazonian rivers. Gradient Forest, redundancy analysis (RDA) and BayPass analyses were used to test for signals of natural selection, and model-based and model-free approaches were used to evaluate neutral population differentiation. After controlling for a signal of neutral hierarchical structure which was consistent with the expectations for a dendritic system, variation in turbidity and pH were key factors contributing to adaptive divergence. Variation in genes involved in acid-sensitive ion transport pathways and light-sensitive photoreceptor pathways was strongly associated with pH and turbidity variability. This study improves our understanding of how natural selection and neutral evolution impact on the distribution of aquatic biodiversity from the understudied and ecologically complex Amazonia.

Machine learning in postgenomic biology and personalized medicine

In recent years Artificial Intelligence in the form of machine learning has been revolutionizing biology, biomedical sciences, and gene-based agricultural technology capabilities. Massive data generated in biological sciences by rapid and deep gene sequencing and protein or other molecular structure determination, on the one hand, requires data analysis capabilities using machine learning that are distinctly different from classical statistical methods; on the other, these large datasets are enabling the adoption of novel data-intensive machine learning algorithms for the solution of biological problems that until recently had relied on mechanistic model-based approaches that are computationally expensive. This review provides a bird's eye view of the applications of machine learning in post-genomic biology. Attempt is also made to indicate as far as possible the areas of research that are poised to make further impacts in these areas, including the importance of explainable artificial intelligence (XAI) in human health. Further contributions of machine learning are expected to transform medicine, public health, agricultural technology, as well as to provide invaluable gene-based guidance for the management of complex environments in this age of global warming.

Population genomics and geographic dispersal in Chagas disease vectors: Landscape drivers and evidence of possible adaptation to the domestic setting

Accurate prediction of vectors dispersal, as well as identification of adaptations that allow blood-feeding vectors to thrive in built environments, are a basis for effective disease control. Here we adopted a landscape genomics approach to assay gene flow, possible local adaptation, and drivers of population structure in Rhodnius ecuadoriensis, an important vector of Chagas disease. We used a reduced-representation sequencing technique (2b-RADseq) to obtain 2,552 SNP markers across 272 R. ecuadoriensis samples from 25 collection sites in southern Ecuador. Evidence of high and directional gene flow between seven wild and domestic population pairs across our study site indicates insecticide-based control will be hindered by repeated re-infestation of houses from the forest. Preliminary genome scans across multiple population pairs revealed shared outlier loci potentially consistent with local adaptation to the domestic setting, which we mapped to genes involved with embryogenesis and saliva production. Landscape genomic models showed elevation is a key barrier to R. ecuadoriensis dispersal. Together our results shed early light on the genomic adaptation in triatomine vectors and facilitate vector control by predicting that spatially-targeted, proactive interventions would be more efficacious than current, reactive approaches.

Re-infestation of recently insecticide-treated houses by wild/secondary triatomine, their potential adaptation to this new environment and capabilities to geographically disperse across multiple human communities jeopardise sustainable Chagas disease control. This is the first study in Chagas disease vectors that identifies genomic regions possibly linked to adaptations to the built environment and describes landscape drivers for accurate prediction of geographic dispersal. We sampled multiple domestic and wild Rhodnius ecuadoriensis population pairs across a mountainous terrain in southern Ecuador. We evidenced that triatomine movement from forest to built enviroments does occur at a high rate. In these highly connected population pairs we detected loci possibly linked to local adaptation among the genomic makers we evaluated and in doing so we pave the way for future triatomine genomic research. We highlighted that current haphazardous vector control in the zone will be hindered by reinfestation of triatomines from the forest. Instead, we recommend frequent and spatially-targeted vector control and provided a landacape genomic model that identifies highly connected and isolated triatomine populations to facilitate efficient vector control.

Adaptive evolution of major histocompatibility complex class I immune genes and disease associations in coastal juvenile sea turtles

Characterizing polymorphism at the major histocompatibility complex (MHC) genes is key to understanding the vertebrate immune response to disease. Despite being globally afflicted by the infectious tumour disease fibropapillomatosis (FP), immunogenetic variation in sea turtles is minimally explored. We sequenced the α 1 peptide-binding region of MHC class I genes (162 bp) from 268 juvenile green (Chelonia mydas) and 88 loggerhead (Caretta caretta) sea turtles in Florida, USA. We recovered extensive variation (116 alleles) and trans-species polymorphism. Supertyping analysis uncovered three functional MHC supertypes corresponding to the three well-supported clades in the phylogeny. We found significant evidence of positive selection at seven amino acid sites in the class I exon. Random forest modelling and risk ratio analysis of Ch. mydas alleles uncovered one allele weakly associated with smooth FP tumour texture, which may be associated with disease outcome. Our study represents the first characterization of MHC class I diversity in Ch. mydas and the largest sample of sea turtles used to date in any study of adaptive genetic variation, revealing tremendous genetic variation and high adaptive potential to viral pathogen threats. The novel associations we identified between MHC diversity and FP outcomes in sea turtles further highlight the importance of evaluating genetic predictors of disease, including MHC and other functional markers.

Development of Novel Management Tools for Phortica variegata (Diptera: Drosophilidae), Vector of the Oriental Eyeworm, Thelazia callipaeda (Spirurida: Thelaziidae), in Europe

Lachryphagous males of Phortica variegata (Fallén, 1823) are gaining increasing attention in Europe, as they act as vectors of the nematode Thelazia callipaeda Railliet & Henry, 1910, causal agent of thelaziosis, an emergent zoonotic disease. Currently, there are no effective control strategies against the vector, and surveillance and monitoring rely on time-consuming and nonselective sampling methods. Our aim was to improve the knowledge about the population dynamics and the chemical ecology of the species. A total of 5,726 P. variegata flies (96.4% males and 3.6% females, mostly gravid) were collected in field experiments during June-September of 2020 in an oak forest in northern Spain. Our results indicate that 1) by means of sweep netting a significantly higher number of captures were found both around the collector´s body and in the air than at ground level; 2) a positive relationship was detected between the abundance of Phortica flies and temperature, with two significant peaks of abundance at 24 and 33°C; 3) the blend of red wine and cider vinegar was the most attractive bait; 4) yellow traps captured fewer flies compared to black and transparent traps; and 5) a significant reduction toward vinegar and wine was detected in presence of the phenolic monoterpenoid carvacrol. In addition, all the males (n = 690) analyzed by both molecular detection and dissection resulted negative for the presence of T. callipaeda larvae. Overall, these findings provide a better understanding of the vector in terms of monitoring and management strategies.

Two decades of association mapping: Insights on disease resistance in major crops

Climate change across the globe has an impact on the occurrence, prevalence, and severity of plant diseases. About 30% of yield losses in major crops are due to plant diseases; emerging diseases are likely to worsen the sustainable production in the coming years. Plant diseases have led to increased hunger and mass migration of human populations in the past, thus a serious threat to global food security. Equipping the modern varieties/hybrids with enhanced genetic resistance is the most economic, sustainable and environmentally friendly solution. Plant geneticists have done tremendous work in identifying stable resistance in primary genepools and many times other than primary genepools to breed resistant varieties in different major crops. Over the last two decades, the availability of crop and pathogen genomes due to advances in next generation sequencing technologies improved our understanding of trait genetics using different approaches. Genome-wide association studies have been effectively used to identify candidate genes and map loci associated with different diseases in crop plants. In this review, we highlight successful examples for the discovery of resistance genes to many important diseases. In addition, major developments in association studies, statistical models and bioinformatic tools that improve the power, resolution and the efficiency of identifying marker-trait associations. Overall this review provides comprehensive insights into the two decades of advances in GWAS studies and discusses the challenges and opportunities this research area provides for breeding resistant varieties.

Time matters: Resilience of a post-disturbance forest landscape

Present-day disturbances are transforming European forest landscapes, and their legacies determine the vulnerability and resilience of the emergent forest generation. To understand these legacy effects, we investigated the resilience of the aboveground forest biomass (B_abg) to a sequence of disturbances affecting the forest in different recovery phases from the initial large-scale impact. We used the model iLand to simulate windthrows that affected 13-24% of the B_abg in a Central European forest landscape. An additional wind event was simulated 20, 40, 60, or 80 years after the initial impact (i.e., sequences of two windthrows were defined). Each windthrow triggered an outbreak of bark beetles that interacted with the recovery processes. We evaluated the resistance of the B_abg to and recovery after the impact. Random Forest models were used to identify factors influencing resilience. We found that B_abg resistance was the lowest 20 years after the initial impact when the increased proportion of emergent wind-exposed forest edges prevailed the disturbance-dampening effect of reduced biomass levels and increased landscape heterogeneity. This forest had a remarkably high recovery rate and reached the pre-disturbance B_abg within 28 years. The forest exhibited a higher resistance and a slower recovery rate in the more advanced recovery phases, reaching the pre-disturbance B_abg within 60-80 years. The recovery was enhanced by higher levels of alpha and beta diversity. Under elevated air temperature, the bark beetle outbreak triggered by windthrow delayed the recovery. However, the positive effect of increased temperature on forest productivity caused the recovery rate to be higher under the warming scenario than under the reference climate. We conclude that resilience is not a static property, but its magnitude and drivers vary in time, depending on vegetation feedbacks, interactions between disturbances, and climate. Understanding these mechanisms is an essential step towards the operationalization of resilience-oriented stewardship.

Predicting physical activity intensity using raw accelerometer signals in manual wheelchair users with spinal cord injury

STUDY DESIGN

Cross-sectional validation study.

OBJECTIVES

The performance of previously published physical activity (PA) intensity cutoff thresholds based on proprietary ActiGraph counts for manual wheelchair users (MWUs) with spinal cord injury (SCI) was initially evaluated using an out-of-sample dataset of 60 individuals with SCI. Two types of PA intensity classification models based on raw accelerometer signals were developed and evaluated.

SETTING

Research institutions in Pittsburgh PA, Birmingham AL, and Bronx NY.

METHODS

Data were collected from 60 MWUs with SCI who followed a structured activity protocol while wearing an ActiGraph activity monitor on their dominant wrist and portable metabolic cart which measured criterion PA intensity. Data was used to assess published models as well as develop and assess custom models using recall, specificity, precision, as well as normalized Mathew's correlation coefficient (nMCC).

RESULTS

All the models performed well for predicting sedentary vs non-sedentary activity, yielding an nMCC of 0.87-0.90. However, all models demonstrated inadequate performance for predicting moderate to vigorous PA (MVPA) with an nMCC of 0.76-0.82.

CONCLUSIONS

The mean absolute deviation (MAD) cutoff threshold yielded the best performance for predicting sedentary vs non-sedentary PA and may be used for tracking daily sedentary activity. None of the models displayed strong performance for MVPA vs non-MVPA. Future studies should investigate combining physiological measures with accelerometry to yield better prediction accuracies for MVPA.

The genetic basis of spatial cognitive variation in a food-caching bird

Spatial cognition is used by most organisms to navigate their environment. Some species rely particularly heavily on specialized spatial cognition to survive, suggesting that a heritable component of cognition may be under natural selection. This idea remains largely untested outside of humans, perhaps because cognition in general is known to be strongly affected by learning and experience.^1-4 We investigated the genetic basis of individual variation in spatial cognition used by non-migratory food-caching birds to recover food stores and survive harsh montane winters. Comparing the genomes of wild, free-living birds ranging from best to worst in their performance on a spatial cognitive task revealed significant associations with genes involved in neuron growth and development and hippocampal function. These results identify candidate genes associated with differences in spatial cognition and provide a critical link connecting individual variation in spatial cognition with natural selection.

Effect of Fire Frequency on the Flammability of Two Mediterranean Pines: Link with Needle Terpene Content

Flammability is a major factor involved in Mediterranean plant evolution that has led to the diversity of fire-related traits according to fire regimes and fire-adaptive strategies. With on-going climate change, new fire regimes are threatening plant species if they do not adapt or acclimate. Studying flammability and terpene content variation according to the different fire frequencies in the recent fire history represents a great challenge to anticipating the flammability of ecosystems in the near future. The flammability of shoots and litter as well as the needle terpene contents of two pine species with different fire adaptive strategies (Pinus halepensis and Pinus sylvestris) were measured according to two fire modalities (0 vs. 1–2 fire events over the last 60 years). Results showed that, regardless of the species and the fuel type, flammability was higher in populations having undergone at least one past fire event even when factors influencing flammability (e.g., structural traits and hydric content) were considered. The terpene content did not vary in P. sylvestris’ needles according to the fire modality, but that of sesqui- and diterpenes was higher in P. halepensis’ needles sampled in the “Fire” modality. In addition, associations made between flammability and terpene content using random forest analyses indicated that the terpene molecules differed between fire modalities for both species and fuel types. The same results were obtained with significant terpenes driving flammability as were highlighted in the PLS analyses, especially for P. halepensis for which enhanced shoot flammability in the “Fire” modality agreed with the adaptive strategy of this species to fire.

Study becomes insight: Ecological learning from machine learning

The ecological and environmental science communities have embraced machine learning (ML) for empirical modelling and prediction. However, going beyond prediction to draw insights into underlying functional relationships between response variables and environmental ‘drivers’ is less straightforward. Deriving ecological insights from fitted ML models requires techniques to extract the ‘learning’ hidden in the ML models.

We revisit the theoretical background and effectiveness of four approaches for deriving insights from ML: ranking independent variable importance (Gini importance, GI; permutation importance, PI; split importance, SI; and conditional permutation importance, CPI), and two approaches for inference of bivariate functional relationships (partial dependence plots, PDP; and accumulated local effect plots, ALE). We also explore the use of a surrogate model for visualization and interpretation of complex multi‐variate relationships between response variables and environmental drivers. We examine the challenges and opportunities for extracting ecological insights with these interpretation approaches. Specifically, we aim to improve interpretation of ML models by investigating how effectiveness relates to (a) interpretation algorithm, (b) sample size and (c) the presence of spurious explanatory variables.

We base the analysis on simulations with known underlying functional relationships between response and predictor variables, with added white noise and the presence of correlated but non‐influential variables. The results indicate that deriving ecological insight is strongly affected by interpretation algorithm and spurious variables, and moderately impacted by sample size. Removing spurious variables improves interpretation of ML models. Meanwhile, increasing sample size has limited value in the presence of spurious variables, but increasing sample size does improves performance once spurious variables are omitted. Among the four ranking methods, SI is slightly more effective than the other methods in the presence of spurious variables, while GI and SI yield higher accuracy when spurious variables are removed. PDP is more effective in retrieving underlying functional relationships than ALE, but its reliability declines sharply in the presence of spurious variables. Visualization and interpretation of the interactive effects of predictors and the response variable can be enhanced using surrogate models, including three‐dimensional visualizations and use of loess planes to represent independent variable effects and interactions.

Machine learning analysts should be aware that including correlated independent variables in ML models with no clear causal relationship to response variables can interfere with ecological inference. When ecological inference is important, ML models should be constructed with independent variables that have clear causal effects on response variables. While interpreting ML models for ecological inference remains challenging, we show that careful choice of interpretation methods, exclusion of spurious variables and adequate sample size can provide more and better opportunities to ‘learn from machine learning’.

Towards fine-scale population stratification modeling based on kernel principal component analysis and random forest

BACKGROUND

Population stratification modeling is essential in Genome-Wide Association Studies.

OBJECTIVE

In this paper, we aim to build a fine-scale population stratification model to efficiently infer individual genetic ancestry.

METHODS

Kernel Principal Component Analysis (PCA) and random forest are adopted to build the population stratification model, together with parameter optimization. We explore different PCA methods, including standard PCA and kernel PCA to extract relevant features from the genotype data that is transformed by vcf2geno, a pipeline from LASER software. These extracted features are fed into a random forest for ensemble learning. Parameter tuning is performed to jointly find the optimal number of principal components, kernel function for PCA and parameters of the random forest.

RESULTS

Experiments based on HGDP dataset show that kernel PCA with Sigmoid function and Gaussian function can achieve higher prediction accuracy than the standard PCA. Compared to standard PCA with the two principal components, the accuracy by using KPCA-Sigmoid with the optimal number of principal components can achieve around 100% and 200% improvement for East Asian and European populations, respectively.

CONCLUSION

With the optimal parameter configuration on both PCA and random forest, our proposed method can infer the individual genetic ancestry more accurately, given their variants.

Genome-Wide Association Studies Reveal Susceptibility Loci for Noninfectious Claw Lesions in Holstein Dairy Cattle

Sole ulcers (SUs) and white line disease (WLD) are two common noninfectious claw lesions (NICL) that arise due to a compromised horn production and are frequent causes of lameness in dairy cattle, imposing welfare and profitability concerns. Low to moderate heritability estimates of SU and WLD susceptibility indicate that genetic selection could reduce their prevalence. To identify the susceptibility loci for SU, WLD, SU and/or WLD, and any type of noninfectious claw lesion, genome-wide association studies (GWAS) were performed using generalized linear mixed model (GLMM) regression, chunk-based association testing (CBAT), and a random forest (RF) approach. Cows from five commercial dairies in California were classified as controls having no lameness records and ≥6 years old (n = 102) or cases having SU (n = 152), WLD (n = 117), SU and/or WLD (SU + WLD, n = 198), or any type of noninfectious claw lesion (n = 217). The top single nucleotide polymorphisms (SNPs) were defined as those passing the Bonferroni-corrected suggestive and significance thresholds in the GLMM analysis or those that a validated RF model considered important. Effects of the top SNPs were quantified using Bayesian estimation. Linkage disequilibrium (LD) blocks defined by the top SNPs were explored for candidate genes and previously identified, functionally relevant quantitative trait loci. The GLMM and CBAT approaches revealed the same regions of association on BTA8 for SU and BTA13 common to WLD, SU + WLD, and NICL. These SNPs had effects significantly different from zero, and the LD blocks they defined explained a significant amount of phenotypic variance for each dataset (6.1-8.1%, p < 0.05), indicating the small but notable contribution of these regions to susceptibility. These regions contained candidate genes involved in wound healing, skin lesions, bone growth and mineralization, adipose tissue, and keratinization. The LD block defined by the most significant SNP on BTA8 for SU included a SNP previously associated with SU. The RF models were overfitted, indicating that the SNP effects were very small, thereby preventing meaningful interpretation of SNPs and any downstream analyses. These findings suggested that variants associated with various physiological systems may contribute to susceptibility for NICL, demonstrating the complexity of genetic predisposition.

Identification and Functional Annotation of Genes Related to Bone Stability in Laying Hens Using Random Forests

Skeletal disorders, including fractures and osteoporosis, in laying hens cause major welfare and economic problems. Although genetics have been shown to play a key role in bone integrity, little is yet known about the underlying genetic architecture of the traits. This study aimed to identify genes associated with bone breaking strength and bone mineral density of the tibiotarsus and the humerus in laying hens. Potentially informative single nucleotide polymorphisms (SNP) were identified using Random Forests classification. We then searched for genes known to be related to bone stability in close proximity to the SNPs and identified 16 potential candidates. Some of them had human orthologues. Based on our findings, we can support the assumption that multiple genes determine bone strength, with each of them having a rather small effect, as illustrated by our SNP effect estimates. Furthermore, the enrichment analysis showed that some of these candidates are involved in metabolic pathways critical for bone integrity. In conclusion, the identified candidates represent genes that may play a role in the bone integrity of chickens. Although further studies are needed to determine causality, the genes reported here are promising in terms of alleviating bone disorders in laying hens.

Correlational selection in the age of genomics

Ecologists and evolutionary biologists are well aware that natural and sexual selection do not operate on traits in isolation, but instead act on combinations of traits. This long-recognized and pervasive phenomenon is known as multivariate selection, or-in the particular case where it favours correlations between interacting traits-correlational selection. Despite broad acknowledgement of correlational selection, the relevant theory has often been overlooked in genomic research. Here, we discuss theory and empirical findings from ecological, quantitative genetic and genomic research, linking key insights from different fields. Correlational selection can operate on both discrete trait combinations and quantitative characters, with profound implications for genomic architecture, linkage, pleiotropy, evolvability, modularity, phenotypic integration and phenotypic plasticity. We synthesize current knowledge and discuss promising research approaches that will enable us to understand how correlational selection shapes genomic architecture, thereby linking quantitative genetic approaches with emerging genomic methods. We suggest that research on correlational selection has great potential to integrate multiple fields in evolutionary biology, including developmental and functional biology, ecology, quantitative genetics, phenotypic polymorphisms, hybrid zones and speciation processes.

Prospects and limitations of genomic offset in conservation management

In nature conservation, there is keen interest in predicting how populations will respond to environmental changes such as climate change. These predictions can help determine whether a population can be self-sustaining under future alterations of its habitat or whether it may require human intervention such as protection, restoration, or assisted migration. An increasingly popular approach in this respect is the concept of genomic offset, which combines genomic and environmental data from different time points and/or locations to assess the degree of possible maladaptation to new environmental conditions. Here, we argue that the concept of genomic offset holds great potential, but an exploration of its risks and limitations is needed to use it for recommendations in conservation or assisted migration. After briefly describing the concept, we list important issues to consider (e.g., statistical frameworks, population genetic structure, migration, independent evidence) when using genomic offset or developing these methods further. We conclude that genomic offset is an area of development that still lacks some important features and should be used in combination with other approaches to inform conservation measures.

Eye-color and Type-2 diabetes phenotype prediction from genotype data using deep learning methods

Background

Genotype–phenotype predictions are of great importance in genetics. These predictions can help to find genetic mutations causing variations in human beings. There are many approaches for finding the association which can be broadly categorized into two classes, statistical techniques, and machine learning. Statistical techniques are good for finding the actual SNPs causing variation where Machine Learning techniques are good where we just want to classify the people into different categories. In this article, we examined the Eye-color and Type-2 diabetes phenotype. The proposed technique is a hybrid approach consisting of some parts from statistical techniques and remaining from Machine learning.

Results

The main dataset for Eye-color phenotype consists of 806 people. 404 people have Blue-Green eyes where 402 people have Brown eyes. After preprocessing we generated 8 different datasets, containing different numbers of SNPs, using the mutation difference and thresholding at individual SNP. We calculated three types of mutation at each SNP no mutation, partial mutation, and full mutation. After that data is transformed for machine learning algorithms. We used about 9 classifiers, RandomForest, Extreme Gradient boosting, ANN, LSTM, GRU, BILSTM, 1DCNN, ensembles of ANN, and ensembles of LSTM which gave the best accuracy of 0.91, 0.9286, 0.945, 0.94, 0.94, 0.92, 0.95, and 0.96% respectively. Stacked ensembles of LSTM outperformed other algorithms for 1560 SNPs with an overall accuracy of 0.96, AUC = 0.98 for brown eyes, and AUC = 0.97 for Blue-Green eyes. The main dataset for Type-2 diabetes consists of 107 people where 30 people are classified as cases and 74 people as controls. We used different linear threshold to find the optimal number of SNPs for classification. The final model gave an accuracy of 0.97%.

Conclusion

Genotype–phenotype predictions are very useful especially in forensic. These predictions can help to identify SNP variant association with traits and diseases. Given more datasets, machine learning model predictions can be increased. Moreover, the non-linearity in the Machine learning model and the combination of SNPs Mutations while training the model increases the prediction. We considered binary classification problems but the proposed approach can be extended to multi-class classification.

Evaluating the performance of machine learning methods and variable selection methods for predicting difficult-to-measure traits in Holstein dairy cattle using milk infrared spectral data

Fourier-transform infrared (FTIR) spectroscopy is a powerful high-throughput phenotyping tool for predicting traits that are expensive and difficult to measure in dairy cattle. Calibration equations are often developed using standard methods, such as partial least squares (PLS) regression. Methods that employ penalization, rank-reduction, and variable selection, as well as being able to model the nonlinear relations between phenotype and FTIR, might offer improvements in predictive ability and model robustness. This study aimed to compare the predictive ability of 2 machine learning methods, namely random forest (RF) and gradient boosting machine (GBM), and penalized regression against PLS regression for predicting 3 phenotypes differing in terms of biological meaning and relationships with milk composition (i.e., phenotypes measurable directly and not directly in milk, reflecting different biological processes which can be captured using milk spectra) in Holstein-Friesian cattle under 2 cross-validation scenarios. The data set comprised phenotypic information from 471 Holstein-Friesian cows, and 3 target phenotypes were evaluated: (1) body condition score (BCS), (2) blood β-hydroxybutyrate (BHB, mmol/L), and (3) κ-casein expressed as a percentage of nitrogen (κ-CN, % N). The data set was split considering 2 cross-validation scenarios: samples-out random in which the population was randomly split into 10-folds (8-folds for training and 1-fold for validation and testing); and herd/date-out in which the population was randomly assigned to training (70% herd), validation (10%), and testing (20% herd) based on the herd and date in which the samples were collected. The random grid search was performed using the training subset for the hyperparameter optimization and the validation set was used for the generalization of prediction error. The trained model was then used to assess the final prediction in the testing subset. The grid search for penalized regression evidenced that the elastic net (EN) was the best regularization with increase in predictive ability of 5%. The performance of PLS (standard model) was compared against 2 machine learning techniques and penalized regression using 2 cross-validation scenarios. Machine learning methods showed a greater predictive ability for BCS (0.63 for GBM and 0.61 for RF), BHB (0.80 for GBM and 0.79 for RF), and κ-CN (0.81 for GBM and 0.80 for RF) in samples-out cross-validation. Considering a herd/date-out cross-validation these values were 0.58 (GBM and RF) for BCS, 0.73 (GBM and RF) for BHB, and 0.77 (GBM and RF) for κ-CN. The GBM model tended to outperform other methods in predictive ability around 4%, 1%, and 7% for EN, RF, and PLS, respectively. The prediction accuracies of the GBM and RF models were similar, and differed statistically from the PLS model in samples-out random cross-validation. Although, machine learning techniques outperformed PLS in herd/date-out cross-validation, no significant differences were observed in terms of predictive ability due to the large standard deviation observed for predictions. Overall, GBM achieved the highest accuracy of FTIR-based prediction of the different phenotypic traits across the cross-validation scenarios. These results indicate that GBM is a promising method for obtaining more accurate FTIR-based predictions for different phenotypes in dairy cattle.

Genotype by environment interactions in coral bleaching

Climate-driven reef decline has prompted the development of next-generation coral conservation strategies, many of which hinge on the movement of adaptive variation across genetic and environmental gradients. This process is limited by our understanding of how genetic and genotypic drivers of coral bleaching will manifest in different environmental conditions. We reciprocally transplanted 10 genotypes of Acropora cervicornis across eight sites along a 60 km span of the Florida Reef Tract and documented significant genotype × environment interactions in bleaching response during the severe 2015 bleaching event. Performance relative to site mean was significantly different between genotypes and can be mostly explained by ensemble models of correlations with genetic markers. The high explanatory power was driven by significant enrichment of loci associated DNA repair, cell signalling and apoptosis. No genotypes performed above (or below) bleaching average at all sites, so genomic predictors can provide practitioners with 'confidence intervals' about the chance of success in novel habitats. These data have important implications for assisted gene flow and managed relocation, and their integration with traditional active restoration.

Development of a predictive model for integrated medical and long-term care resource consumption based on health behaviour: application of healthcare big data of patients with circulatory diseases

BACKGROUND

Medical costs and the burden associated with cardiovascular disease are on the rise. Therefore, to improve the overall economy and quality assessment of the healthcare system, we developed a predictive model of integrated healthcare resource consumption (Adherence Score for Healthcare Resource Outcome, ASHRO) that incorporates patient health behaviours, and examined its association with clinical outcomes.

METHODS

This study used information from a large-scale database on health insurance claims, long-term care insurance, and health check-ups. Participants comprised patients who received inpatient medical care for diseases of the circulatory system (ICD-10 codes I00-I99). The predictive model used broadly defined composite adherence as the explanatory variable and medical and long-term care costs as the objective variable. Predictive models used random forest learning (AI: artificial intelligence) to adjust for predictors, and multiple regression analysis to construct ASHRO scores. The ability of discrimination and calibration of the prediction model were evaluated using the area under the curve and the Hosmer-Lemeshow test. We compared the overall mortality of the two ASHRO 50% cut-off groups adjusted for clinical risk factors by propensity score matching over a 48-month follow-up period.

RESULTS

Overall, 48,456 patients were discharged from the hospital with cardiovascular disease (mean age, 68.3 ± 9.9 years; male, 61.9%). The broad adherence score classification, adjusted as an index of the predictive model by machine learning, was an index of eight: secondary prevention, rehabilitation intensity, guidance, proportion of days covered, overlapping outpatient visits/clinical laboratory and physiological tests, medical attendance, and generic drug rate. Multiple regression analysis showed an overall coefficient of determination of 0.313 (p < 0.001). Logistic regression analysis with cut-off values of 50% and 25%/75% for medical and long-term care costs showed that the overall coefficient of determination was statistically significant (p < 0.001). The score of ASHRO was associated with the incidence of all deaths between the two 50% cut-off groups (2% vs. 7%; p < 0.001).

CONCLUSIONS

ASHRO accurately predicted future integrated healthcare resource consumption and was associated with clinical outcomes. It can be a valuable tool for evaluating the economic usefulness of individual adherence behaviours and optimising clinical outcomes.

Distribution, sources, and decomposition of soil organic matter along a salinity gradient in estuarine wetlands characterized by C:N ratio, δ13 C-δ15 N, and lignin biomarker

Despite increasing recognition of the critical role of coastal wetlands in mitigating climate change, sea-level rise, and salinity increase, soil organic carbon (SOC) sequestration mechanisms in estuarine wetlands remain poorly understood. Here, we present new results on the source, decomposition, and storage of SOC in estuarine wetlands with four vegetation types, including single Phragmites australis (P, habitat I), a mixture of P. australis and Suaeda salsa (P + S, habitat II), single S. salsa (S, habitat III), and tidal flat (TF, habitat IV) across a salinity gradient. Values of δ¹³ C increased with depth in aerobic soil layers (0-40 cm) but slightly decreased in anaerobic soil layers (40-100 cm). The δ¹⁵ N was significantly enriched in soil organic matter at all depths than in the living plant tissues, indicating a preferential decomposition of ¹⁴ N-enriched organic components. Thus, the kinetic isotope fractionation during microbial degradation and the preferential substrate utilization are the dominant mechanisms in regulating isotopic compositions in aerobic and anaerobic conditions, respectively. Stable isotopic (δ¹³ C and δ¹⁵ N), elemental (C and N), and lignin composition (inherited (Ad/Al)s and C/V) were not completely consistent in reflecting the differences in SOC decomposition or accumulation among four vegetation types, possibly due to differences in litter inputs, root distributions, substrate quality, water-table level, salinity, and microbial community composition/activity. Organic C contents and storage decreased from upstream to downstream, likely due to primarily changes in autochthonous sources (e.g., decreased onsite plant biomass input) and allochthonous materials (e.g., decreased fluvially transported upland river inputs, and increased tidally induced marine algae and phytoplankton). Our results revealed that multiple indicators are essential to unravel the degree of SOC decomposition and accumulation, and a combination of C:N ratios, δ¹³ C, δ¹⁵ N, and lignin biomarker provides a robust approach to decipher the decomposition and source of sedimentary organic matter along the river-estuary-ocean continuum.

Comprehensive Genomic Investigation of Adaptive Mutations Driving the Low-Level Oxacillin Resistance Phenotype in Staphylococcus aureus

Antistaphylococcal penicillins such as oxacillin are the key antibiotics in the treatment of invasive methicillin-susceptible Staphylococcus aureus (MSSA) infections; however, mec gene-independent resistance adaptation can cause treatment failure. Despite its clinical relevance, the basis of this phenomenon remains poorly understood. Here, we investigated the genomic adaptation to oxacillin at an unprecedented scale using a large collection of 503 clinical mec-negative isolates and 30 in vitro-adapted isolates from independent oxacillin exposures. By combining comparative genomics, evolutionary convergence, and genome-wide association analysis, we found 21 genetic loci associated with low-level oxacillin resistance, underscoring the polygenic nature of this phenotype. Evidence of adaptation was particularly strong for the c-di-AMP signal transduction pathways (gdpP and dacA) and in the clpXP chaperone-protease complex. The role of mutations in gdpP in conferring low-level oxacillin resistance was confirmed by allele-swapping experiments. We found that resistance to oxacillin emerges at high frequency in vitro (median, 2.9 × 10^-6; interquartile range [IQR], 1.9 × 10^-6 to 3.9 × 10^-6), which is consistent with a recurrent minimum inhibitory concentration (MIC) increase across the global phylogeny of clinical isolates. Nevertheless, adaptation in clinical isolates appears sporadically, with no stably adapted lineages, suggesting a high fitness cost of resistance, confirmed by growth assessment of mutants in rich media. Our data provide a broader understanding of the emergence and dynamics of oxacillin resistance adaptation in S. aureus and a framework for future surveillance of this clinically important phenomenon.IMPORTANCE The majority of Staphylococcus aureus strains causing human disease are methicillin-susceptible (MSSA) and can be treated with antistaphylococcal penicillins (such as oxacillin). While acquisition of the mec gene represents the main resistance mechanism to oxacillin, S. aureus can acquire low-level resistance through adaptive mutations in other genes. In this study, we used genomic approaches to understand the basis of S. aureus adaption to oxacillin and its dynamic at the population level. By combining a genome analysis of clinical isolates from persistent MSSA infections, in vitro selection of oxacillin resistance, and genome-wide association analysis on a large collection of isolates, we identified 21 genes linked to secondary oxacillin resistance. Adaptive mutations in these genes were easy to select when S. aureus was exposed to oxacillin, but they also came at a substantial cost in terms of bacterial fitness, suggesting that this phenotype emerges preferentially in the setting of sustained antibiotic exposure.

Combining Random Forests and a Signal Detection Method Leads to the Robust Detection of Genotype-Phenotype Associations

Genome wide association studies (GWAS) are a well established methodology to identify genomic variants and genes that are responsible for traits of interest in all branches of the life sciences. Despite the long time this methodology has had to mature the reliable detection of genotype-phenotype associations is still a challenge for many quantitative traits mainly because of the large number of genomic loci with weak individual effects on the trait under investigation. Thus, it can be hypothesized that many genomic variants that have a small, however real, effect remain unnoticed in many GWAS approaches. Here, we propose a two-step procedure to address this problem. In a first step, cubic splines are fitted to the test statistic values and genomic regions with spline-peaks that are higher than expected by chance are considered as quantitative trait loci (QTL). Then the SNPs in these QTLs are prioritized with respect to the strength of their association with the phenotype using a Random Forests approach. As a case study, we apply our procedure to real data sets and find trustworthy numbers of, partially novel, genomic variants and genes involved in various egg quality traits.

Comparing the utility of in vivo transposon mutagenesis approaches in yeast species to infer gene essentiality

In vivo transposon mutagenesis, coupled with deep sequencing, enables large-scale genome-wide mutant screens for genes essential in different growth conditions. We analyzed six large-scale studies performed on haploid strains of three yeast species (Saccharomyces cerevisiae, Schizosaccaromyces pombe, and Candida albicans), each mutagenized with two of three different heterologous transposons (AcDs, Hermes, and PiggyBac). Using a machine-learning approach, we evaluated the ability of the data to predict gene essentiality. Important data features included sufficient numbers and distribution of independent insertion events. All transposons showed some bias in insertion site preference because of jackpot events, and preferences for specific insertion sequences and short-distance vs long-distance insertions. For PiggyBac, a stringent target sequence limited the ability to predict essentiality in genes with few or no target sequences. The machine learning approach also robustly predicted gene function in less well-studied species by leveraging cross-species orthologs. Finally, comparisons of isogenic diploid versus haploid S. cerevisiae isolates identified several genes that are haplo-insufficient, while most essential genes, as expected, were recessive. We provide recommendations for the choice of transposons and the inference of gene essentiality in genome-wide studies of eukaryotic haploid microbes such as yeasts, including species that have been less amenable to classical genetic studies.

Identification of Age-Specific and Common Key Regulatory Mechanisms Governing Eggshell Strength in Chicken Using Random Forests

In today's chicken egg industry, maintaining the strength of eggshells in longer laying cycles is pivotal for improving the persistency of egg laying. Eggshell development and mineralization underlie a complex regulatory interplay of various proteins and signaling cascades involving multiple organ systems. Understanding the regulatory mechanisms influencing this dynamic trait over time is imperative, yet scarce. To investigate the temporal changes in the signaling cascades, we considered eggshell strength at two different time points during the egg production cycle and studied the genotype-phenotype associations by employing the Random Forests algorithm on chicken genotypic data. For the analysis of corresponding genes, we adopted a well established systems biology approach to delineate gene regulatory pathways and master regulators underlying this important trait. Our results indicate that, while some of the master regulators (Slc22a1 and Sox11) and pathways are common at different laying stages of chicken, others (e.g., Scn11a, St8sia2, or the TGF- β pathway) represent age-specific functions. Overall, our results provide: (i) significant insights into age-specific and common molecular mechanisms underlying the regulation of eggshell strength; and (ii) new breeding targets to improve the eggshell quality during the later stages of the chicken production cycle.

The worldwide impact of urbanisation on avian functional diversity

Urbanisation is driving rapid declines in species richness and abundance worldwide, but the general implications for ecosystem function and services remain poorly understood. Here, we integrate global data on bird communities with comprehensive information on traits associated with ecological processes to show that assemblages in highly urbanised environments have substantially different functional composition and 20% less functional diversity on average than surrounding natural habitats. These changes occur without significant decreases in functional dissimilarity between species; instead, they are caused by a decrease in species richness and abundance evenness, leading to declines in functional redundancy. The reconfiguration and decline of native functional diversity in cities are not compensated by the presence of exotic species but are less severe under moderate levels of urbanisation. Thus, urbanisation has substantial negative impacts on functional diversity, potentially resulting in impaired provision of ecosystem services, but these impacts can be reduced by less intensive urbanisation practices.

Estimating Forest Aboveground Carbon Storage in Hang-Jia-Hu Using Landsat TM/OLI Data and Random Forest Model

Dynamic monitoring of carbon storage in forests resources is important for tracking ecosystem functionalities and climate change impacts. In this study, we used multi-year Landsat data combined with a Random Forest (RF) algorithm to estimate the forest aboveground carbon (AGC) in a forest area in China (Hang-Jia-Hu) and analyzed its spatiotemporal changes during the past two decades. Maximum likelihood classification was applied to make land-use maps. Remote sensing variables, such as the spectral band, vegetation indices, and derived texture features, were extracted from 20 Landsat TM and OLI images over five different years (2000, 2004, 2010, 2015, and 2018). These variables were subsequently selected according to their importance and subsequently used in the RF algorithm to build an estimation model of forest AGC. The results showed the following: (1) Verification of classification results showed maximum likelihood can extract land information effectively. Our land cover classification yielded overall accuracies between 86.86% and 89.47%. (2) Additionally, our RF models showed good performance in predicting forest AGC, with R2 from 0.65 to 0.73 in the training and testing phase and a RMSE range between 3.18 and 6.66 Mg/ha. RMSEr in the testing phase ranged from 20.27 to 22.27 with a low model error. (3) The estimation results indicated that forest AGC in the past two decades increased with density at 10.14 Mg/ha, 21.63 Mg/ha, 26.39 Mg/ha, 29.25 Mg/ha, and 44.59 Mg/ha in 2000, 2004, 2010, 2015, and 2018. The total forest AGC storage had a growth rate of 285%. (4) Our study showed that, although forest area decreased in the study area during the time period under study, the total forest AGC increased due to an increment in forest AGC density. However, such an effect is overridden in the vicinity of cities by intense urbanization and the loss of forest covers. Our study demonstrated that the combined use of remote sensing data and machine learning techniques can improve our ability to track the forest changes in support of regional natural resource management practices.

The use of classification and regression algorithms using the random forests method with presence-only data to model species' distribution

Random forests (RF) is a powerful species distribution model (SDM) algorithm. This ensemble model by default can produce categorical and numerical species distribution maps based on its classification tree (CT) and regression tree (RT) algorithms, respectively. The CT algorithm can also produce numerical predictions (class probability). Here, we present a detailed procedure involving the use of the CT and RT algorithms using the RF method with presence-only data to model the distribution of species. CT and RT are used to generate numerical prediction maps, and then numerical predictions are converted to binary predictions through objective threshold-setting methods. We also applied simple methods to deal with collinearity of predictor variables and spatial autocorrelation of species occurrence data. A geographically stratified sampling method was employed for generating pseudo-absences. The detailed procedural framework is meant to be a generic method to be applied to virtually any SDM prediction question using presence-only data.

•

How to use RF as a standard method for generic species distributions with presence-only data

•

How to choose RF (CT or RT) methods for the distribution modeling of species

•

A general and detailed procedure for any SDM prediction question.