Evolutionarily informed deep learning methods for predicting relative transcript abundance from DNA sequence

Identification, Design, and Application of Noncoding Cis-Regulatory Elements

Cis-regulatory elements (CREs) play a pivotal role in orchestrating interactions with trans-regulatory factors such as transcription factors, RNA-binding proteins, and noncoding RNAs. These interactions are fundamental to the molecular architecture underpinning complex and diverse biological functions in living organisms, facilitating a myriad of sophisticated and dynamic processes. The rapid advancement in the identification and characterization of these regulatory elements has been marked by initiatives such as the Encyclopedia of DNA Elements (ENCODE) project, which represents a significant milestone in the field. Concurrently, the development of CRE detection technologies, exemplified by massively parallel reporter assays, has progressed at an impressive pace, providing powerful tools for CRE discovery. The exponential growth of multimodal functional genomic data has necessitated the application of advanced analytical methods. Deep learning algorithms, particularly large language models, have emerged as invaluable tools for deconstructing the intricate nucleotide sequences governing CRE function. These advancements facilitate precise predictions of CRE activity and enable the de novo design of CREs. A deeper understanding of CRE operational dynamics is crucial for harnessing their versatile regulatory properties. Such insights are instrumental in refining gene therapy techniques, enhancing the efficacy of selective breeding programs, pushing the boundaries of genetic innovation, and opening new possibilities in microbial synthetic biology.

Applying interpretable machine learning in computational biology-pitfalls, recommendations and opportunities for new developments

Recent advances in machine learning have enabled the development of next-generation predictive models for complex computational biology problems, thereby spurring the use of interpretable machine learning (IML) to unveil biological insights. However, guidelines for using IML in computational biology are generally underdeveloped. We provide an overview of IML methods and evaluation techniques and discuss common pitfalls encountered when applying IML methods to computational biology problems. We also highlight open questions, especially in the era of large language models, and call for collaboration between IML and computational biology researchers.

Deep Learning Sequence Models for Transcriptional Regulation

Deciphering the regulatory code of gene expression and interpreting the transcriptional effects of genome variation are critical challenges in human genetics. Modern experimental technologies have resulted in an abundance of data, enabling the development of sequence-based deep learning models that link patterns embedded in DNA to the biochemical and regulatory properties contributing to transcriptional regulation, including modeling epigenetic marks, 3D genome organization, and gene expression, with tissue and cell-type specificity. Such methods can predict the functional consequences of any noncoding variant in the human genome, even rare or never-before-observed variants, and systematically characterize their consequences beyond what is tractable from experiments or quantitative genetics studies alone. Recently, the development and application of interpretability approaches have led to the identification of key sequence patterns contributing to the predicted tasks, providing insights into the underlying biological mechanisms learned and revealing opportunities for improvement in future models.

Arabidopsis and maize terminator strength is determined by GC content, polyadenylation motifs and cleavage probability

The 3' end of a gene, often called a terminator, modulates mRNA stability, localization, translation, and polyadenylation. Here, we adapted Plant STARR-seq, a massively parallel reporter assay, to measure the activity of over 50,000 terminators from the plants Arabidopsis thaliana and Zea mays. We characterize thousands of plant terminators, including many that outperform bacterial terminators commonly used in plants. Terminator activity is species-specific, differing in tobacco leaf and maize protoplast assays. While recapitulating known biology, our results reveal the relative contributions of polyadenylation motifs to terminator strength. We built a computational model to predict terminator strength and used it to conduct in silico evolution that generated optimized synthetic terminators. Additionally, we discover alternative polyadenylation sites across tens of thousands of terminators; however, the strongest terminators tend to have a dominant cleavage site. Our results establish features of plant terminator function and identify strong naturally occurring and synthetic terminators.

A foundational large language model for edible plant genomes

Significant progress has been made in the field of plant genomics, as demonstrated by the increased use of high-throughput methodologies that enable the characterization of multiple genome-wide molecular phenotypes. These findings have provided valuable insights into plant traits and their underlying genetic mechanisms, particularly in model plant species. Nonetheless, effectively leveraging them to make accurate predictions represents a critical step in crop genomic improvement. We present AgroNT, a foundational large language model trained on genomes from 48 plant species with a predominant focus on crop species. We show that AgroNT can obtain state-of-the-art predictions for regulatory annotations, promoter/terminator strength, tissue-specific gene expression, and prioritize functional variants. We conduct a large-scale in silico saturation mutagenesis analysis on cassava to evaluate the regulatory impact of over 10 million mutations and provide their predicted effects as a resource for variant characterization. Finally, we propose the use of the diverse datasets compiled here as the Plants Genomic Benchmark (PGB), providing a comprehensive benchmark for deep learning-based methods in plant genomic research. The pre-trained AgroNT model is publicly available on HuggingFace at https://huggingface.co/InstaDeepAI/agro-nucleotide-transformer-1b  for future research purposes.

Modeling 0.6 million genes for the rational design of functional cis-regulatory variants and de novo design of cis-regulatory sequences

Rational design of plant cis-regulatory DNA sequences without expert intervention or prior domain knowledge is still a daunting task. Here, we developed PhytoExpr, a deep learning framework capable of predicting both mRNA abundance and plant species using the proximal regulatory sequence as the sole input. PhytoExpr was trained over 17 species representative of major clades of the plant kingdom to enhance its generalizability. Via input perturbation, quantitative functional annotation of the input sequence was achieved at single-nucleotide resolution, revealing an abundance of predicted high-impact nucleotides in conserved noncoding sequences and transcription factor binding sites. Evaluation of maize HapMap3 single-nucleotide polymorphisms (SNPs) by PhytoExpr demonstrates an enrichment of predicted high-impact SNPs in cis-eQTL. Additionally, we provided two algorithms that harnessed the power of PhytoExpr in designing functional cis-regulatory variants, and de novo creation of species-specific cis-regulatory sequences through in silico evolution of random DNA sequences. Our model represents a general and robust approach for functional variant discovery in population genetics and rational design of regulatory sequences for genome editing and synthetic biology.

DeepCBA: A deep learning framework for gene expression prediction in maize based on DNA sequences and chromatin interactions

Chromatin interactions create spatial proximity between distal regulatory elements and target genes in the genome, which has an important impact on gene expression, transcriptional regulation, and phenotypic traits. To date, several methods have been developed for predicting gene expression. However, existing methods do not take into consideration the effect of chromatin interactions on target gene expression, thus potentially reducing the accuracy of gene expression prediction and mining of important regulatory elements. In this study, we developed a highly accurate deep learning-based gene expression prediction model (DeepCBA) based on maize chromatin interaction data. Compared with existing models, DeepCBA exhibits higher accuracy in expression classification and expression value prediction. The average Pearson correlation coefficients (PCCs) for predicting gene expression using gene promoter proximal interactions, proximal-distal interactions, and both proximal and distal interactions were 0.818, 0.625, and 0.929, respectively, representing an increase of 0.357, 0.16, and 0.469 over the PCCs obtained with traditional methods that use only gene proximal sequences. Some important motifs were identified through DeepCBA; they were enriched in open chromatin regions and expression quantitative trait loci and showed clear tissue specificity. Importantly, experimental results for the maize flowering-related gene ZmRap2.7 and the tillering-related gene ZmTb1 demonstrated the feasibility of DeepCBA for exploration of regulatory elements that affect gene expression. Moreover, promoter editing and verification of two reported genes (ZmCLE7 and ZmVTE4) demonstrated the utility of DeepCBA for the precise design of gene expression and even for future intelligent breeding. DeepCBA is available at http://www.deepcba.com/ or http://124.220.197.196/.

Large language models in plant biology

Large language models (LLMs), such as ChatGPT, have taken the world by storm. However, LLMs are not limited to human language and can be used to analyze sequential data, such as DNA, protein, and gene expression. The resulting foundation models can be repurposed to identify the complex patterns within the data, resulting in powerful, multipurpose prediction tools able to predict the state of cellular systems. This review outlines the different types of LLMs and showcases their recent uses in biology. Since LLMs have not yet been embraced by the plant community, we also cover how these models can be deployed for the plant kingdom.

ctGAN: combined transformation of gene expression and survival data with generative adversarial network

Recent studies have extensively used deep learning algorithms to analyze gene expression to predict disease diagnosis, treatment effectiveness, and survival outcomes. Survival analysis studies on diseases with high mortality rates, such as cancer, are indispensable. However, deep learning models are plagued by overfitting owing to the limited sample size relative to the large number of genes. Consequently, the latest style-transfer deep generative models have been implemented to generate gene expression data. However, these models are limited in their applicability for clinical purposes because they generate only transcriptomic data. Therefore, this study proposes ctGAN, which enables the combined transformation of gene expression and survival data using a generative adversarial network (GAN). ctGAN improves survival analysis by augmenting data through style transformations between breast cancer and 11 other cancer types. We evaluated the concordance index (C-index) enhancements compared with previous models to demonstrate its superiority. Performance improvements were observed in nine of the 11 cancer types. Moreover, ctGAN outperformed previous models in seven out of the 11 cancer types, with colon adenocarcinoma (COAD) exhibiting the most significant improvement (median C-index increase of ~15.70%). Furthermore, integrating the generated COAD enhanced the log-rank p-value (0.041) compared with using only the real COAD (p-value = 0.797). Based on the data distribution, we demonstrated that the model generated highly plausible data. In clustering evaluation, ctGAN exhibited the highest performance in most cases (89.62%). These findings suggest that ctGAN can be meaningfully utilized to predict disease progression and select personalized treatments in the medical field.

Genotype sampling for deep-learning assisted experimental mapping of a combinatorially complete fitness landscape

MOTIVATION

Experimental characterization of fitness landscapes, which map genotypes onto fitness, is important for both evolutionary biology and protein engineering. It faces a fundamental obstacle in the astronomical number of genotypes whose fitness needs to be measured for any one protein. Deep learning may help to predict the fitness of many genotypes from a smaller neural network training sample of genotypes with experimentally measured fitness. Here I use a recently published experimentally mapped fitness landscape of more than 260 000 protein genotypes to ask how such sampling is best performed.

RESULTS

I show that multilayer perceptrons, recurrent neural networks, convolutional networks, and transformers, can explain more than 90% of fitness variance in the data. In addition, 90% of this performance is reached with a training sample comprising merely ≈103 sequences. Generalization to unseen test data is best when training data is sampled randomly and uniformly, or sampled to minimize the number of synonymous sequences. In contrast, sampling to maximize sequence diversity or codon usage bias reduces performance substantially. These observations hold for more than one network architecture. Simple sampling strategies may perform best when training deep learning neural networks to map fitness landscapes from experimental data.

AVAILABILITY AND IMPLEMENTATION

The fitness landscape data analyzed here is publicly available as described previously (Papkou et al. 2023). All code used to analyze this landscape is publicly available at https://github.com/andreas-wagner-uzh/fitness_landscape_sampling.

Why aquatic scientists should use sulfur stable isotope ratios (ẟ34S) more often

Over the last few decades, measurements of light stable isotope ratios have been increasingly used to answer questions across physiology, biology, ecology, and archaeology. The vast majority analyse carbon (δ13C) and nitrogen (δ15N) stable isotopes as the 'default' isotopes, omitting sulfur (δ34S) due to time, cost, or perceived lack of benefits and instrumentation capabilities. Using just carbon and nitrogen isotopic ratios can produce results that are inconclusive, uncertain, or in the worst cases, even misleading, especially for scientists that are new to the use and interpretation of stable isotope data. Using sulfur isotope values more regularly has the potential to mitigate these issues, especially given recent advancements that have lowered measurement barriers. Here we provide a review documenting case studies with real-world data, re-analysing different biological topics (i.e. niche, physiology, diet, movement and bioarchaeology) with and without sulfur isotopes to highlight the various strengths of this stable isotope for various applications. We also include a preliminary meta-analysis of the trophic discrimination factor (TDF) for sulfur isotopes, which suggest small (mean -0.4 ± 1.7 ‰ SD) but taxa-dependent mean trophic discrimination. Each case study demonstrates how the exclusion of sulfur comes at the detriment of the results, often leading to very different outputs, or missing valuable discoveries entirely. Given that studies relying on carbon and nitrogen stable isotopes currently underpin most of our understanding of various ecological processes, this has concerning implications. Collectively, these examples strongly suggest that researchers planning to use carbon and nitrogen stable isotopes for their research should incorporate sulfur where possible, and that the new 'default' isotope systems for aquatic science should now be carbon, nitrogen, and sulfur.

Predicting gene expression responses to environment in Arabidopsis thaliana using natural variation in DNA sequence

The evolution of gene expression responses are a critical component of adaptation to variable environments. Predicting how DNA sequence influences expression is challenging because the genotype to phenotype map is not well resolved for cis regulatory elements, transcription factor binding, regulatory interactions, and epigenetic features, not to mention how these factors respond to environment. We tested if flexible machine learning models could learn some of the underlying cis- regulatory genotype to phenotype map. We tested this approach using cold-responsive transcriptome profiles in 5 diverse Arabidopsis thaliana accessions. We first tested for evidence that cis regulation plays a role in environmental response, finding 14 and 15 motifs that were significantly enriched within the up- and down-stream regions of cold-responsive differentially regulated genes (DEGs). We next applied convolutional neural networks (CNNs), which learn de novo cis- regulatory motifs in DNA sequences to predict expression response to environment. We found that CNNs predicted differential expression with moderate accuracy, with evidence that predictions were hindered by biological complexity of regulation and the large potential regulatory code. Overall, DEGs between specific environments can be predicted based on variation in cis- regulatory sequences, although more information needs to be incorporated and better models may be required.

AI can empower agriculture for global food security: challenges and prospects in developing nations

Food and nutrition are a steadfast essential to all living organisms. With specific reference to humans, the sufficient and efficient supply of food is a challenge as the world population continues to grow. Artificial Intelligence (AI) could be identified as a plausible technology in this 5th industrial revolution in bringing us closer to achieving zero hunger by 2030-Goal 2 of the United Nations Sustainable Development Goals (UNSDG). This goal cannot be achieved unless the digital divide among developed and underdeveloped countries is addressed. Nevertheless, developing and underdeveloped regions fall behind in economic resources; however, they harbor untapped potential to effectively address the impending demands posed by the soaring world population. Therefore, this study explores the in-depth potential of AI in the agriculture sector for developing and under-developed countries. Similarly, it aims to emphasize the proven efficiency and spin-off applications of AI in the advancement of agriculture. Currently, AI is being utilized in various spheres of agriculture, including but not limited to crop surveillance, irrigation management, disease identification, fertilization practices, task automation, image manipulation, data processing, yield forecasting, supply chain optimization, implementation of decision support system (DSS), weed control, and the enhancement of resource utilization. Whereas AI supports food safety and security by ensuring higher crop yields that are acquired by harnessing the potential of multi-temporal remote sensing (RS) techniques to accurately discern diverse crop phenotypes, monitor land cover dynamics, assess variations in soil organic matter, predict soil moisture levels, conduct plant biomass modeling, and enable comprehensive crop monitoring. The present study identifies various challenges, including financial, infrastructure, experts, data availability, customization, regulatory framework, cultural norms and attitudes, access to market, and interdisciplinary collaboration, in the adoption of AI for developing nations with their subsequent remedies. The identification of challenges and opportunities in the implementation of AI could ignite further research and actions in these regions; thereby supporting sustainable development.

Deep learning the cis-regulatory code for gene expression in selected model plants

Elucidating the relationship between non-coding regulatory element sequences and gene expression is crucial for understanding gene regulation and genetic variation. We explored this link with the training of interpretable deep learning models predicting gene expression profiles from gene flanking regions of the plant species Arabidopsis thaliana, Solanum lycopersicum, Sorghum bicolor, and Zea mays. With over 80% accuracy, our models enabled predictive feature selection, highlighting e.g. the significant role of UTR regions in determining gene expression levels. The models demonstrated remarkable cross-species performance, effectively identifying both conserved and species-specific regulatory sequence features and their predictive power for gene expression. We illustrated the application of our approach by revealing causal links between genetic variation and gene expression changes across fourteen tomato genomes. Lastly, our models efficiently predicted genotype-specific expression of key functional gene groups, exemplified by underscoring known phenotypic and metabolic differences between Solanum lycopersicum and its wild, drought-resistant relative, Solanum pennellii.

Arabidopsis and Maize Terminator Strength is Determined by GC Content, Polyadenylation Motifs and Cleavage Probability

The 3' end of a gene, often called a terminator, modulates mRNA stability, localization, translation, and polyadenylation. Here, we adapted Plant STARR-seq, a massively parallel reporter assay, to measure the activity of over 50,000 terminators from the plants Arabidopsis thaliana and Zea mays. We characterize thousands of plant terminators, including many that outperform bacterial terminators commonly used in plants. Terminator activity is species-specific, differing in tobacco leaf and maize protoplast assays. While recapitulating known biology, our results reveal the relative contributions of polyadenylation motifs to terminator strength. We built a computational model to predict terminator strength and used it to conduct in silico evolution that generated optimized synthetic terminators. Additionally, we discover alternative polyadenylation sites across tens of thousands of terminators; however, the strongest terminators tend to have a dominant cleavage site. Our results establish features of plant terminator function and identify strong naturally occurring and synthetic terminators.

Characterization and optimization of 5´ untranslated region containing poly-adenine tracts in Kluyveromyces marxianus using machine-learning model

BACKGROUND

The 5´ untranslated region (5´ UTR) plays a key role in regulating translation efficiency and mRNA stability, making it a favored target in genetic engineering and synthetic biology. A common feature found in the 5´ UTR is the poly-adenine (poly(A)) tract. However, the effect of 5´ UTR poly(A) on protein production remains controversial. Machine-learning models are powerful tools for explaining the complex contributions of features, but models incorporating features of 5´ UTR poly(A) are currently lacking. Thus, our goal is to construct such a model, using natural 5´ UTRs from Kluyveromyces marxianus, a promising cell factory for producing heterologous proteins.

RESULTS

We constructed a mini-library consisting of 207 5´ UTRs harboring poly(A) and 34 5´ UTRs without poly(A) from K. marxianus. The effects of each 5´ UTR on the production of a GFP reporter were evaluated individually in vivo, and the resulting protein abundance spanned an approximately 450-fold range throughout. The data were used to train a multi-layer perceptron neural network (MLP-NN) model that incorporated the length and position of poly(A) as features. The model exhibited good performance in predicting protein abundance (average R2 = 0.7290). The model suggests that the length of poly(A) is negatively correlated with protein production, whereas poly(A) located between 10 and 30 nt upstream of the start codon (AUG) exhibits a weak positive effect on protein abundance. Using the model as guidance, the deletion or reduction of poly(A) upstream of 30 nt preceding AUG tended to improve the production of GFP and a feruloyl esterase. Deletions of poly(A) showed inconsistent effects on mRNA levels, suggesting that poly(A) represses protein production either with or without reducing mRNA levels.

CONCLUSION

The effects of poly(A) on protein production depend on its length and position. Integrating poly(A) features into machine-learning models improves simulation accuracy. Deleting or reducing poly(A) upstream of 30 nt preceding AUG tends to enhance protein production. This optimization strategy can be applied to enhance the yield of K. marxianus and other microbial cell factories.

Topological data analysis reveals a core gene expression backbone that defines form and function across flowering plants

Since they emerged approximately 125 million years ago, flowering plants have evolved to dominate the terrestrial landscape and survive in the most inhospitable environments on earth. At their core, these adaptations have been shaped by changes in numerous, interconnected pathways and genes that collectively give rise to emergent biological phenomena. Linking gene expression to morphological outcomes remains a grand challenge in biology, and new approaches are needed to begin to address this gap. Here, we implemented topological data analysis (TDA) to summarize the high dimensionality and noisiness of gene expression data using lens functions that delineate plant tissue and stress responses. Using this framework, we created a topological representation of the shape of gene expression across plant evolution, development, and environment for the phylogenetically diverse flowering plants. The TDA-based Mapper graphs form a well-defined gradient of tissues from leaves to seeds, or from healthy to stressed samples, depending on the lens function. This suggests that there are distinct and conserved expression patterns across angiosperms that delineate different tissue types or responses to biotic and abiotic stresses. Genes that correlate with the tissue lens function are enriched in central processes such as photosynthetic, growth and development, housekeeping, or stress responses. Together, our results highlight the power of TDA for analyzing complex biological data and reveal a core expression backbone that defines plant form and function.

Predicting transcriptional responses to heat and drought stress from genomic features using a machine learning approach in rice

Plants have evolved various mechanisms to adapt to adverse environmental stresses, such as the modulation of gene expression. Expression of stress-responsive genes is controlled by specific regulators, including transcription factors (TFs), that bind to sequence-specific binding sites, representing key components of cis-regulatory elements and regulatory networks. Our understanding of the underlying regulatory code remains, however, incomplete. Recent studies have shown that, by training machine learning (ML) algorithms on genomic sequence features, it is possible to predict which genes will transcriptionally respond to a specific stress. By identifying the most important features for gene expression prediction, these trained ML models allow, in theory, to further elucidate the regulatory code underlying the transcriptional response to abiotic stress. Here, we trained random forest ML models to predict gene expression in rice (Oryza sativa) in response to heat or drought stress. Apart from thoroughly assessing model performance and robustness across various input training data, the importance of promoter and gene body sequence features to train ML models was evaluated. The use of enriched promoter oligomers, complementing known TF binding sites, allowed us to gain novel insights in DNA motifs contributing to the stress regulatory code. By comparing genomic feature importance scores for drought and heat stress over time, general and stress-specific genomic features contributing to the performance of the learned models and their temporal variation were identified. This study provides a solid foundation to build and interpret ML models accurately predicting transcriptional responses and enables novel insights in biological sequence features that are important for abiotic stress responses.

DeeP4med: deep learning for P4 medicine to predict normal and cancer transcriptome in multiple human tissues

BACKGROUND

P4 medicine (predict, prevent, personalize, and participate) is a new approach to diagnosing and predicting diseases on a patient-by-patient basis. For the prevention and treatment of diseases, prediction plays a fundamental role. One of the intelligent strategies is the design of deep learning models that can predict the state of the disease using gene expression data.

RESULTS

We create an autoencoder deep learning model called DeeP4med, including a Classifier and a Transferor that predicts cancer's gene expression (mRNA) matrix from its matched normal sample and vice versa. The range of the F1 score of the model, depending on tissue type in the Classifier, is from 0.935 to 0.999 and in Transferor from 0.944 to 0.999. The accuracy of DeeP4med for tissue and disease classification was 0.986 and 0.992, respectively, which performed better compared to seven classic machine learning models (Support Vector Classifier, Logistic Regression, Linear Discriminant Analysis, Naive Bayes, Decision Tree, Random Forest, K Nearest Neighbors).

CONCLUSIONS

Based on the idea of DeeP4med, by having the gene expression matrix of a normal tissue, we can predict its tumor gene expression matrix and, in this way, find effective genes in transforming a normal tissue into a tumor tissue. Results of Differentially Expressed Genes (DEGs) and enrichment analysis on the predicted matrices for 13 types of cancer showed a good correlation with the literature and biological databases. This led that by using the gene expression matrix, to train the model with features of each person in a normal and cancer state, this model could predict diagnosis based on gene expression data from healthy tissue and be used to identify possible therapeutic interventions for those patients.

Recognizing the power of machine learning and other computational methods to accelerate progress in small molecule targeting of RNA

RNA structures regulate a wide range of processes in biology and disease, yet small molecule chemical probes or drugs that can modulate these functions are rare. Machine learning and other computational methods are well poised to fill gaps in knowledge and overcome the inherent challenges in RNA targeting, such as the dynamic nature of RNA and the difficulty of obtaining RNA high-resolution structures. Successful tools to date include principal component analysis, linear discriminate analysis, k-nearest neighbor, artificial neural networks, multiple linear regression, and many others. Employment of these tools has revealed critical factors for selective recognition in RNA:small molecule complexes, predictable differences in RNA- and protein-binding ligands, and quantitative structure activity relationships that allow the rational design of small molecules for a given RNA target. Herein we present our perspective on the value of using machine learning and other computation methods to advance RNA:small molecule targeting, including select examples and their validation as well as necessary and promising future directions that will be key to accelerate discoveries in this important field.

Computational Methods Summarizing Mutational Patterns in Cancer: Promise and Limitations for Clinical Applications

Since the rise of next-generation sequencing technologies, the catalogue of mutations in cancer has been continuously expanding. To address the complexity of the cancer-genomic landscape and extract meaningful insights, numerous computational approaches have been developed over the last two decades. In this review, we survey the current leading computational methods to derive intricate mutational patterns in the context of clinical relevance. We begin with mutation signatures, explaining first how mutation signatures were developed and then examining the utility of studies using mutation signatures to correlate environmental effects on the cancer genome. Next, we examine current clinical research that employs mutation signatures and discuss the potential use cases and challenges of mutation signatures in clinical decision-making. We then examine computational studies developing tools to investigate complex patterns of mutations beyond the context of mutational signatures. We survey methods to identify cancer-driver genes, from single-driver studies to pathway and network analyses. In addition, we review methods inferring complex combinations of mutations for clinical tasks and using mutations integrated with multi-omics data to better predict cancer phenotypes. We examine the use of these tools for either discovery or prediction, including prediction of tumor origin, treatment outcomes, prognosis, and cancer typing. We further discuss the main limitations preventing widespread clinical integration of computational tools for the diagnosis and treatment of cancer. We end by proposing solutions to address these challenges using recent advances in machine learning.

Charting plant gene functions in the multi-omics and single-cell era

Despite the increased access to high-quality plant genome sequences, the set of genes with a known function remains far from complete. With the advent of novel bulk and single-cell omics profiling methods, we are entering a new era where advanced and highly integrative functional annotation strategies are being developed to elucidate the functions of all plant genes. Here, we review different multi-omics approaches to improve functional and regulatory gene characterization and highlight the power of machine learning and network biology to fully exploit the complementary information embedded in different omics layers. Finally, we discuss the potential of emerging single-cell methods and algorithms to further increase the resolution, allowing generation of functional insights about plant biology.

Designing artificial synthetic promoters for accurate, smart, and versatile gene expression in plants

With the development of high-throughput biology techniques and artificial intelligence, it has become increasingly feasible to design and construct artificial biological parts, modules, circuits, and even whole systems. To overcome the limitations of native promoters in controlling gene expression, artificial promoter design aims to synthesize short, inducible, and conditionally controlled promoters to coordinate the expression of multiple genes in diverse plant metabolic and signaling pathways. Synthetic promoters are versatile and can drive gene expression accurately with smart responses; they show potential for enhancing desirable traits in crops, thereby improving crop yield, nutritional quality, and food security. This review first illustrates the importance of synthetic promoters, then introduces promoter architecture and thoroughly summarizes advances in synthetic promoter construction. Restrictions to the development of synthetic promoters and future applications of such promoters in synthetic plant biology and crop improvement are also discussed.

Obtaining genetics insights from deep learning via explainable artificial intelligence

Artificial intelligence (AI) models based on deep learning now represent the state of the art for making functional predictions in genomics research. However, the underlying basis on which predictive models make such predictions is often unknown. For genomics researchers, this missing explanatory information would frequently be of greater value than the predictions themselves, as it can enable new insights into genetic processes. We review progress in the emerging area of explainable AI (xAI), a field with the potential to empower life science researchers to gain mechanistic insights into complex deep learning models. We discuss and categorize approaches for model interpretation, including an intuitive understanding of how each approach works and their underlying assumptions and limitations in the context of typical high-throughput biological datasets.

Deep learning in regulatory genomics: from identification to design

Genomics and deep learning are a natural match since both are data-driven fields. Regulatory genomics refers to functional noncoding DNA regulating gene expression. In recent years, deep learning applications on regulatory genomics have achieved remarkable advances so-much-so that it has revolutionized the rules of the game of the computational methods in this field. Here, we review two emerging trends: (i) the modeling of very long input sequence (up to 200 kb), which requires self-matched modularization of model architecture; (ii) on the balance of model predictability and model interpretability because the latter is more able to meet biological demands. Finally, we discuss how to employ these two routes to design synthetic regulatory DNA, as a promising strategy for optimizing crop agronomic properties.

Crop genomic selection with deep learning and environmental data: A survey

Machine learning techniques for crop genomic selections, especially for single-environment plants, are well-developed. These machine learning models, which use dense genome-wide markers to predict phenotype, routinely perform well on single-environment datasets, especially for complex traits affected by multiple markers. On the other hand, machine learning models for predicting crop phenotype, especially deep learning models, using datasets that span different environmental conditions, have only recently emerged. Models that can accept heterogeneous data sources, such as temperature, soil conditions and precipitation, are natural choices for modeling GxE in multi-environment prediction. Here, we review emerging deep learning techniques that incorporate environmental data directly into genomic selection models.

A Transferable Machine Learning Framework for Predicting Transcriptional Responses of Genes Across Species

Leveraging existing resources in studied species to predict gene functions has the potential to rapidly expand understanding of annotated genes in other, less well-studied, species with assembled genomes. However, orthology is not a reliable predictor for the transcriptional responses of genes to stress. Machine learning methods can quantitatively estimate expression patterns and gene functions using known annotations and collections of features describing each gene. In this chapter, we describe a supervised machine learning framework to predict stress-responsive genes across species using only features derived from nucleotide sequences, using the example of cold stress-responsive genes in different Panicoid grass species.

A survey on gene expression data analysis using deep learning methods for cancer diagnosis

Gene Expression Data is the biological data to extract meaningful hidden information from the gene dataset. This gene information is used for disease diagnosis especially in cancer treatment based on the variations in gene expression levels. DNA microarray is an efficient method for gene expression classification and prediction of cancer disease for specific types of cancer. Due to the abundance of computing power, deep learning (DL) has become a widespread technique in the healthcare sector. The gene expression dataset has a limited number of samples but a large number of features. Data augmentation is needed for gene expression datasets to overcome the dimensionality problem in gene data. It is a technique to generating the synthetic samples to increase the diversity of data. Deep learning methods are designed to learn and extract the features that come from the raw input data in the form of multidimensional arrays. This paper reviews the existing research in deep learning techniques like Feed Forward Neural Network (FFN), Convolutional Neural Network (CNN), Autoencoder (AE) and Recurrent Neural Network (RNN) for the classification and prediction of cancer disease and its types through gene expression data analysis.

Toward learning the principles of plant gene regulation

Advanced machine learning (ML) algorithms produce highly accurate models of gene expression, uncovering novel regulatory features in nucleotide sequences involving multiple cis-regulatory regions across whole genes and structural properties. These broaden our understanding of gene regulation and point to new principles to test and adopt in the field of plant science.

Comprehensive transcriptional variability analysis reveals gene networks regulating seed oil content of Brassica napus

BACKGROUND

Regulation of gene expression plays an essential role in controlling the phenotypes of plants. Brassica napus (B. napus) is an important source for the vegetable oil in the world, and the seed oil content is an important trait of B. napus.

RESULTS

We perform a comprehensive analysis of the transcriptional variability in the seeds of B. napus at two developmental stages, 20 and 40 days after flowering (DAF). We detect 53,759 and 53,550 independent expression quantitative trait loci (eQTLs) for 79,605 and 76,713 expressed genes at 20 and 40 DAF, respectively. Among them, the local eQTLs are mapped to the adjacent genes more frequently. The adjacent gene pairs are regulated by local eQTLs with the same open chromatin state and show a stronger mode of expression piggybacking. Inter-subgenomic analysis indicates that there is a feedback regulation for the homoeologous gene pairs to maintain partial expression dosage. We also identify 141 eQTL hotspots and find that hotspot87-88 co-localizes with a QTL for the seed oil content. To further resolve the regulatory network of this eQTL hotspot, we construct the XGBoost model using 856 RNA-seq datasets and the Basenji model using 59 ATAC-seq datasets. Using these two models, we predict the mechanisms affecting the seed oil content regulated by hotspot87-88 and experimentally validate that the transcription factors, NAC13 and SCL31, positively regulate the seed oil content.

CONCLUSIONS

We comprehensively characterize the gene regulatory features in the seeds of B. napus and reveal the gene networks regulating the seed oil content of B. napus.

Smart breeding approaches in post-genomics era for developing climate-resilient food crops

Improving the crop traits is highly required for the development of superior crop varieties to deal with climate change and the associated abiotic and biotic stress challenges. Climate change-driven global warming can trigger higher insect pest pressures and plant diseases thus affecting crop production sternly. The traits controlling genes for stress or disease tolerance are economically imperative in crop plants. In this scenario, the extensive exploration of available wild, resistant or susceptible germplasms and unraveling the genetic diversity remains vital for breeding programs. The dawn of next-generation sequencing technologies and omics approaches has accelerated plant breeding by providing the genome sequences and transcriptomes of several plants. The availability of decoded plant genomes offers an opportunity at a glance to identify candidate genes, quantitative trait loci (QTLs), molecular markers, and genome-wide association studies that can potentially aid in high throughput marker-assisted breeding. In recent years genomics is coupled with marker-assisted breeding to unravel the mechanisms to harness better better crop yield and quality. In this review, we discuss the aspects of marker-assisted breeding and recent perspectives of breeding approaches in the era of genomics, bioinformatics, high-tech phonemics, genome editing, and new plant breeding technologies for crop improvement. In nutshell, the smart breeding toolkit in the post-genomics era can steadily help in developing climate-smart future food crops.

Modeling chromatin state from sequence across angiosperms using recurrent convolutional neural networks

Accessible chromatin regions are critical components of gene regulation but modeling them directly from sequence remains challenging, especially within plants, whose mechanisms of chromatin remodeling are less understood than in animals. We trained an existing deep-learning architecture, DanQ, on data from 12 angiosperm species to predict the chromatin accessibility in leaf of sequence windows within and across species. We also trained DanQ on DNA methylation data from 10 angiosperms because unmethylated regions have been shown to overlap significantly with ACRs in some plants. The across-species models have comparable or even superior performance to a model trained within species, suggesting strong conservation of chromatin mechanisms across angiosperms. Testing a maize (Zea mays L.) held-out model on a multi-tissue chromatin accessibility panel revealed our models are best at predicting constitutively accessible chromatin regions, with diminishing performance as cell-type specificity increases. Using a combination of interpretation methods, we ranked JASPAR motifs by their importance to each model and saw that the TCP and AP2/ERF transcription factor (TF) families consistently ranked highly. We embedded the top three JASPAR motifs for each model at all possible positions on both strands in our sequence window and observed position- and strand-specific patterns in their importance to the model. With our publicly available across-species 'a2z' model it is now feasible to predict the chromatin accessibility and methylation landscape of any angiosperm genome.

Using Artificial Neural Networks to Model Errors in Biochemical Manipulation of DNA Molecules

In recent years, the non-biological applications of DNA molecules have made considerable progress; most of these applications were performed in vitro, involving biochemical operations such as synthesis, amplification and sequencing. Because errors may occur with specific sequence patterns or experimental instruments, these biochemical operations are not completely reliable. Modeling errors in these biochemical procedures is an interesting research topic. For example, researchers have proposed several methods to avoid the known vulnerable sequence patterns in the study of storing binary information in DNA molecules. However, there are few end-to-end methods to evaluate these biochemical errors with regard to the DNA sequences. In this article, based on the data generated by a DNA storage research, we use artificial neural networks to predict whether a DNA sequence tends to cause errors in biochemical operations. Through comparative experiments and hyperparameter optimization, we analyze the known and potential problems in the research process. As a result, an end-to-end method to model the biochemical errors of DNA molecules in vitro through a computer system is proposed.

Predicting which genes will respond to transcription factor perturbations

The ability to predict which genes will respond to the perturbation of a transcription factor serves as a benchmark for our systems-level understanding of transcriptional regulatory networks. In previous work, machine learning models have been trained to predict static gene expression levels in a biological sample by using data from the same or similar samples, including data on their transcription factor binding locations, histone marks, or DNA sequence. We report on a different challenge-training machine learning models to predict which genes will respond to the perturbation of a transcription factor without using any data from the perturbed cells. We find that existing transcription factor location data (ChIP-seq) from human cells have very little detectable utility for predicting which genes will respond to perturbation of a transcription factor. Features of genes, including their preperturbation expression level and expression variation, are very useful for predicting responses to perturbation of any transcription factor. This shows that some genes are poised to respond to transcription factor perturbations and others are resistant, shedding light on why it has been so difficult to predict responses from binding locations. Certain histone marks, including H3K4me1 and H3K4me3, have some predictive power when located downstream of the transcription start site. However, the predictive power of histone marks is much less than that of gene expression level and expression variation. Sequence-based or epigenetic properties of genes strongly influence their tendency to respond to direct transcription factor perturbations, partially explaining the oft-noted difficulty of predicting responsiveness from transcription factor binding location data. These molecular features are largely reflected in and summarized by the gene's expression level and expression variation. Code is available at https://github.com/BrentLab/TFPertRespExplainer.

Predicting Tissue-Specific mRNA and Protein Abundance in Maize: A Machine Learning Approach

Machine learning and modeling approaches have been used to classify protein sequences for a broad set of tasks including predicting protein function, structure, expression, and localization. Some recent studies have successfully predicted whether a given gene is expressed as mRNA or even translated to proteins potentially, but given that not all genes are expressed in every condition and tissue, the challenge remains to predict condition-specific expression. To address this gap, we developed a machine learning approach to predict tissue-specific gene expression across 23 different tissues in maize, solely based on DNA promoter and protein sequences. For class labels, we defined high and low expression levels for mRNA and protein abundance and optimized classifiers by systematically exploring various methods and combinations of k-mer sequences in a two-phase approach. In the first phase, we developed Markov model classifiers for each tissue and built a feature vector based on the predictions. In the second phase, the feature vector was used as an input to a Bayesian network for final classification. Our results show that these methods can achieve high classification accuracy of up to 95% for predicting gene expression for individual tissues. By relying on sequence alone, our method works in settings where costly experimental data are unavailable and reveals useful insights into the functional, evolutionary, and regulatory characteristics of genes.

Genome-wide cis-decoding for expression design in tomato using cistrome data and explainable deep learning

In the evolutionary history of plants, variation in cis-regulatory elements (CREs) resulting in diversification of gene expression has played a central role in driving the evolution of lineage-specific traits. However, it is difficult to predict expression behaviors from CRE patterns to properly harness them, mainly because the biological processes are complex. In this study, we used cistrome datasets and explainable convolutional neural network (CNN) frameworks to predict genome-wide expression patterns in tomato (Solanum lycopersicum) fruit from the DNA sequences in gene regulatory regions. By fixing the effects of trans-acting factors using single cell-type spatiotemporal transcriptome data for the response variables, we developed a prediction model for crucial expression patterns in the initiation of tomato fruit ripening. Feature visualization of the CNNs identified nucleotide residues critical to the objective expression pattern in each gene, and their effects were validated experimentally in ripening tomato fruit. This cis-decoding framework will not only contribute to the understanding of the regulatory networks derived from CREs and transcription factor interactions, but also provides a flexible means of designing alleles for optimized expression.

Machine learning: its challenges and opportunities in plant system biology

Sequencing technologies are evolving at a rapid pace, enabling the generation of massive amounts of data in multiple dimensions (e.g., genomics, epigenomics, transcriptomic, metabolomics, proteomics, and single-cell omics) in plants. To provide comprehensive insights into the complexity of plant biological systems, it is important to integrate different omics datasets. Although recent advances in computational analytical pipelines have enabled efficient and high-quality exploration and exploitation of single omics data, the integration of multidimensional, heterogenous, and large datasets (i.e., multi-omics) remains a challenge. In this regard, machine learning (ML) offers promising approaches to integrate large datasets and to recognize fine-grained patterns and relationships. Nevertheless, they require rigorous optimizations to process multi-omics-derived datasets. In this review, we discuss the main concepts of machine learning as well as the key challenges and solutions related to the big data derived from plant system biology. We also provide in-depth insight into the principles of data integration using ML, as well as challenges and opportunities in different contexts including multi-omics, single-cell omics, protein function, and protein-protein interaction. KEY POINTS: • The key challenges and solutions related to the big data derived from plant system biology have been highlighted. • Different methods of data integration have been discussed. • Challenges and opportunities of the application of machine learning in plant system biology have been highlighted and discussed.

Variation in upstream open reading frames contributes to allelic diversity in maize protein abundance

SignificanceProteins are the machinery which execute essential cellular functions. However, measuring their abundance within an organism can be difficult and resource-intensive. Cells use a variety of mechanisms to control protein synthesis from mRNA, including short open reading frames (uORFs) that lie upstream of the main coding sequence. Ribosomes can preferentially translate uORFs instead of the main coding sequence, leading to reduced translation of the main protein. In this study, we show that uORF sequence variation between individuals can lead to different rates of protein translation and thus variable protein abundances. We also demonstrate that natural variation in uORFs occurs frequently and can be linked to whole-plant phenotypes, indicating that uORF sequence variation likely contributes to plant adaptation.

Interpretation of convolutional neural networks reveals crucial sequence features involving in transcription during fiber development

BACKGROUND

Upland cotton provides the most natural fiber in the world. During fiber development, the quality and yield of fiber were influenced by gene transcription. Revealing sequence features related to transcription has a profound impact on cotton molecular breeding. We applied convolutional neural networks to predict gene expression status based on the sequences of gene transcription start regions. After that, a gradient-based interpretation and an N-adjusted kernel transformation were implemented to extract sequence features contributing to transcription.

RESULTS

Our models had approximate 80% accuracies, and the area under the receiver operating characteristic curve reached over 0.85. Gradient-based interpretation revealed 5' untranslated region contributed to gene transcription. Furthermore, 6 DOF binding motifs and 4 transcription activator binding motifs were obtained by N-adjusted kernel-motif transformation from models in three developmental stages. Apart from 10 general motifs, 3 DOF5.1 genes were also detected. In silico analysis about these motifs' binding proteins implied their potential functions in fiber formation. Besides, we also found some novel motifs in plants as important sequence features for transcription.

CONCLUSIONS

In conclusion, the N-adjusted kernel transformation method could interpret convolutional neural networks and reveal important sequence features related to transcription during fiber development. Potential functions of motifs interpreted from convolutional neural networks could be validated by further wet-lab experiments and applied in cotton molecular breeding.

Prediction of conserved and variable heat and cold stress response in maize using cis-regulatory information

Changes in gene expression are important for responses to abiotic stress. Transcriptome profiling of heat- or cold-stressed maize genotypes identifies many changes in transcript abundance. We used comparisons of expression responses in multiple genotypes to identify alleles with variable responses to heat or cold stress and to distinguish examples of cis- or trans-regulatory variation for stress-responsive expression changes. We used motifs enriched near the transcription start sites (TSSs) for thermal stress-responsive genes to develop predictive models of gene expression responses. Prediction accuracies can be improved by focusing only on motifs within unmethylated regions near the TSS and vary for genes with different dynamic responses to stress. Models trained on expression responses in a single genotype and promoter sequences provided lower performance when applied to other genotypes but this could be improved by using models trained on data from all three genotypes tested. The analysis of genes with cis-regulatory variation provides evidence for structural variants that result in presence/absence of transcription factor binding sites in creating variable responses. This study provides insights into cis-regulatory motifs for heat- and cold-responsive gene expression and defines a framework for developing models to predict expression responses across multiple genotypes.

Prediction of whole-cell transcriptional response with machine learning

MOTIVATION

Applications in synthetic and systems biology can benefit from measuring whole-cell response to biochemical perturbations. Execution of experiments to cover all possible combinations of perturbations is infeasible. In this paper, we present the host response model (HRM), a machine learning approach that maps response of single perturbations to transcriptional response of the combination of perturbations.

RESULTS

The HRM combines high-throughput sequencing with machine learning to infer links between experimental context, prior knowledge of cell regulatory networks, and RNASeq data to predict a gene's dysregulation. We find that the HRM can predict the directionality of dysregulation to a combination of inducers with an accuracy of >90% using data from single inducers. We further find that the use of prior, known cell regulatory networks doubles the predictive performance of the HRM (an R2 from 0.3 to 0.65). The model was validated in two organisms, Escherichia coli and Bacillus subtilis, using new experiments conducted after training. Finally, while the HRM is trained with gene expression data, the direct prediction of differential expression makes it possible to also conduct enrichment analyses using its predictions. We show that the HRM can accurately classify >95% of the pathway regulations. The HRM reduces the number of RNASeq experiments needed as responses can be tested in silico prior to the experiment.

AVAILABILITY AND IMPLEMENTATION

The HRM software and tutorial are available at https://github.com/sd2e/CDM and the configurable differential expression analysis tools and tutorials are available at https://github.com/SD2E/omics_tools.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Boosting the power of transcriptomics by developing an efficient gene expression profiling approach

Recent advances in plant genomics are scaling up gene expression profiling from the individual level to the population level, making transcriptomics a more powerful tool while deciphering the genome function. This study developed an efficient 3'RNA-seq method, Simplified Poly(A) Anchored Sequencing (SiPAS), to perform large-scale experiments of gene expression quantification. Aside from being cost-effective, by conducting a comprehensive performance assessment of SiPAS in hexaploid wheat, we demonstrated that SiPAS is highly sensitive, accurate, and reproducible while quantifying gene expression. Our method is anticipated to boost studies of population transcriptomics in plants and improve our understanding of genome biology.

Quantitative Extraction and Evaluation of Tomato Fruit Phenotypes Based on Image Recognition

Tomato fruit phenotypes are important agronomic traits in tomato breeding as a reference index. The traditional measurement methods based on manual observation, however, limit the high-throughput data collection of tomato fruit morphologies. In this study, fruits of 10 different tomato cultivars with considerable differences in fruit color, size, and other morphological characters were selected as samples. Constant illumination condition was applied to take images of the selected tomato fruit samples. Based on image recognition, automated methods for measuring color and size indicators of tomato fruit phenotypes were proposed. A deep learning model based on Mask Region-Convolutional Neural Network (R-CNN) was trained and tested to analyze the internal structure indicators of tomato fruit. The results revealed that the combined use of these methods can extract various important fruit phenotypes of tomato, including fruit color, horizontal and vertical diameters, top and navel angles, locule number, and pericarp thickness, automatically. Considering several corrections of missing and wrong segmentation cases in practice, the average precision of the deep learning model is more than 0.95 in practice. This suggests a promising locule segmentation and counting performance. Vertical/horizontal ratio (fruit shape index) and locule area proportion were also calculated based on the data collected here. The measurement precision was comparable to manual operation, and the measurement efficiency was highly improved. The results of this study will provide a new option for more accurate and efficient tomato fruit phenotyping, which can effectively avoid artificial error and increase the support efficiency of relevant data in the future breeding work of tomato and other fruit crops.

Bluster or Lustre: Can AI Improve Crops and Plant Health?

In a changing climate where future food security is a growing concern, researchers are exploring new methods and technologies in the effort to meet ambitious crop yield targets. The application of Artificial Intelligence (AI) including Machine Learning (ML) methods in this area has been proposed as a potential mechanism to support this. This review explores current research in the area to convey the state-of-the-art as to how AI/ML have been used to advance research, gain insights, and generally enable progress in this area. We address the question-Can AI improve crops and plant health? We further discriminate the bluster from the lustre by identifying the key challenges that AI has been shown to address, balanced with the potential issues with its usage, and the key requisites for its success. Overall, we hope to raise awareness and, as a result, promote usage, of AI related approaches where they can have appropriate impact to improve practices in agricultural and plant sciences.

The arches and spandrels of maize domestication, adaptation, and improvement

People living in the Balsas River basin in southwest México domesticated maize from the bushy grass teosinte. Nine thousand years later, in 2021, Ms. Deb Haaland - a member of the Pueblo of Laguna tribe of New Mexico - wore a dress adorned with a cornstalk when she was sworn in as the Secretary of Interior of the United States of America. This choice of garment highlights the importance of the coevolution of maize and the farmers who, through careful selection over thousands of years, domesticated maize and adapted the physiology and shoot architecture of maize to fit local environments and growth habits. Some traits such as tillering were directly selected on (arches), and others such as tassel size are the by-products (spandrels) of maize evolution. Here, we review current knowledge of the underlying cellular, developmental, physiological, and metabolic processes that were selected by farmers and breeders, which have positioned maize as a top global staple crop.

Metabolomics and complementary techniques to investigate the plant phytochemical cosmos

Covering: up to 2021Plants and their associated microbial communities are known to produce millions of metabolites, a majority of which are still not characterized and are speculated to possess novel bioactive properties. In addition to their role in plant physiology, these metabolites are also relevant as existing and next-generation medicine candidates. Elucidation of the plant metabolite diversity is thus valuable for the successful exploitation of natural resources for humankind. Herein, we present a comprehensive review on recent metabolomics approaches to illuminate molecular networks in plants, including chemical isolation and enzymatic production as well as the modern metabolomics approaches such as stable isotope labeling, ultrahigh-resolution mass spectrometry, metabolome imaging (spatial metabolomics), single-cell analysis, cheminformatics, and computational mass spectrometry. Mass spectrometry-based strategies to characterize plant metabolomes through metabolite identification and annotation are described in detail. We also highlight the use of phytochemical genomics to mine genes associated with specialized metabolites' biosynthesis. Understanding the metabolic diversity through biotechnological advances is fundamental to elucidate the functions of the plant-derived specialized metabolome.

Haplotype associated RNA expression (HARE) improves prediction of complex traits in maize

Genomic prediction typically relies on associations between single-site polymorphisms and traits of interest. This representation of genomic variability has been successful for predicting many complex traits. However, it usually cannot capture the combination of alleles in haplotypes and it has generated little insight about the biological function of polymorphisms. Here we present a novel and cost-effective method for imputing cis haplotype associated RNA expression (HARE), studied their transferability across tissues, and evaluated genomic prediction models within and across populations. HARE focuses on tightly linked cis acting causal variants in the immediate vicinity of the gene, while excluding trans effects from diffusion and metabolism. Therefore, HARE estimates were more transferrable across different tissues and populations compared to measured transcript expression. We also showed that HARE estimates captured one-third of the variation in gene expression. HARE estimates were used in genomic prediction models evaluated within and across two diverse maize panels–a diverse association panel (Goodman Association panel) and a large half-sib panel (Nested Association Mapping panel)–for predicting 26 complex traits. HARE resulted in up to 15% higher prediction accuracy than control approaches that preserved haplotype structure, suggesting that HARE carried functional information in addition to information about haplotype structure. The largest increase was observed when the model was trained in the Nested Association Mapping panel and tested in the Goodman Association panel. Additionally, HARE yielded higher within-population prediction accuracy as compared to measured expression values. The accuracy achieved by measured expression was variable across tissues, whereas accuracy by HARE was more stable across tissues. Therefore, imputing RNA expression of genes by haplotype is stable, cost-effective, and transferable across populations.

Genomic marker data is widely used in the prediction of many traits. However, prediction has been primarily carried out within populations and without explicit modeling of RNA or protein expression. In this study, we explored the prediction of field traits within and across populations using estimated RNA expression attributable to only the DNA sequence around a gene. We showed that the estimated RNA expression was more transferable across populations and tissues than measured RNA expression. We improved prediction of field traits up to 15% using estimated gene expression as compared to observed expression or gene sequence alone. Overall, these findings indicate that structural and functional information in the gene sequence is highly transferable.

Transcriptional regulation in plants: Using omics data to crack the cis-regulatory code

Innovative omics technologies, advanced bioinformatics, and machine learning methods are rapidly becoming integral tools for plant functional genomics, with tremendous recent advances made in this field. In transcriptional regulation, an initial lag in the accumulation of plant omics data relative to that of animals stimulated the development of computational methods capable of extracting maximum information from the available data sets. Recent comprehensive studies of transcription factor-binding profiles in Arabidopsis and maize and the accumulation of uniformly processed omics data in public databases have brought plant biologists into the big leagues, with many cutting-edge methods available. Here, we summarize the state-of-the-art bioinformatics approaches used to predict or infer the cis-regulatory code behind transcriptional gene regulation, focusing on their plant research applications.

A deep learning approach to automate whole-genome prediction of diverse epigenomic modifications in plants

Epigenetic modifications function in gene transcription, RNA metabolism, and other biological processes. However, multiple factors currently limit the scientific utility of epigenomic datasets generated for plants. Here, using deep-learning approaches, we developed a Smart Model for Epigenetics in Plants (SMEP) to predict six types of epigenomic modifications: DNA 5-methylcytosine (5mC) and N6-methyladenosine (6mA) methylation, RNA N6-methyladenosine (m⁶ A) methylation, and three types of histone modification. Using the datasets from the japonica rice Nipponbare, SMEP achieved 95% prediction accuracy for 6mA, and also achieved around 80% for 5mC, m⁶ A, and the three types of histone modification based on the 10-fold cross-validation. Additionally, > 95% of the 6mA peaks detected after a heat-shock treatment were predicted. We also successfully applied the SMEP for examining epigenomic modifications in indica rice 93-11 and even the B73 maize line. Taken together, we show that the deep-learning-enabled SMEP can reliably mine epigenomic datasets from diverse plants to yield actionable insights about epigenomic sites. Thus, our work opens new avenues for the application of predictive tools to facilitate functional research, and will almost certainly increase the efficiency of genome engineering efforts.

Machine Learning for Causal Inference in Biological Networks: Perspectives of This Challenge

Most machine learning-based methods predict outcomes rather than understanding causality. Machine learning methods have been proved to be efficient in finding correlations in data, but unskilful to determine causation. This issue severely limits the applicability of machine learning methods to infer the causal relationships between the entities of a biological network, and more in general of any dynamical system, such as medical intervention strategies and clinical outcomes system, that is representable as a network. From the perspective of those who want to use the results of network inference not only to understand the mechanisms underlying the dynamics, but also to understand how the network reacts to external stimuli (e. g. environmental factors, therapeutic treatments), tools that can understand the causal relationships between data are highly demanded. Given the increasing popularity of machine learning techniques in computational biology and the recent literature proposing the use of machine learning techniques for the inference of biological networks, we would like to present the challenges that mathematics and computer science research faces in generalising machine learning to an approach capable of understanding causal relationships, and the prospects that achieving this will open up for the medical application domains of systems biology, the main paradigm of which is precisely network biology at any physical scale.