Assessing genomic prediction accuracy for Holstein sires using bootstrap aggregation sampling and leave-one-out cross validation

Causal effect of the age at first birth with depression: a mendelian randomization study

BACKGROUND

This study aimed to explore the causal relationship between age at first birth (AFB) and depression.

METHODS

Using the univariable Mendelian randomization (UVMR) and multivariable Mendelian randomization (MVMR) methods to examine the potential correlation between age at first birth (AFB) and major depressive disorder and postpartum depression. A public database was used to obtain the genome-wide association studies (GWAS) summary data. We put inverse-variance-weighted (IVW) as the primary method in Mendelian randomization (MR) analysis and used sensitivity analysis to confirm the robustness of our result.

RESULTS

We found a significant causal association between AFB and major depressive disorder by using the IVW algorithm (odd ratio [OR] 0.826; 95% confidence interval [CI] 0.793 - 0.861; P = 4.51 × 10- 20). MR-Egger, weighted median, simple mode and weighted mode method concluded the same result (P < 0.05). During the sensitivity analysis, the heterogeneity test (Q-value = 55.061, df = 48, P = 2.81 × 10- 01, I2 = 12.82%) and the leave-one-out plot analysis confirmed the stability of the results. The outcomes of the pleiotropy test (MR-Egger intercept = 8.932 × 10- 3. SE = 6.909 × 10- 3. P = 2.02 × 10- 01) and MR_PRESSO global test (P = 2.03 × 10- 01) indicated there is no pleiotropy.

CONCLUSION

There is solid evidence that a higher age at first birth is associated with a lower risk of major depressive disorder.

Causal relationship between atrial fibrillation and stroke risk: a Mendelian randomization

OBJECTIVES

This study aimed to investigate the causal relationship between Atrial Fibrillation (AF) and the risk of Stroke using a Mendelian randomization (MR) approach.

METHODS

A two-sample MR analysis was conducted using publicly available genome-wide association study (GWAS) summary statistics data. In this analysis, genetic variants associated with AF were used as instrumental variables to estimate the causal effect. The inverse-variance weighted (IVW) method, weighted median estimator, and MR-Egger regression were employed for estimation. Additionally, sensitivity analysis was performed using the leave-one-out method.

RESULTS

The analysis included 87 single nucleotide polymorphisms (SNPs) associated with AF. The results from the IVW method indicated a positive association between genetic predisposition to AF and the risk of stroke (OR 1.002, 95 % CI 1.001-1.003, P < 0.001). The weighted median and MR-Egger methods showed consistent results (weighted median: OR 1.001, 95 % CI 1.000-1.002, P = 0.034; MR-Egger: OR 1.001, 95 % CI 1.000-1.003, P = 0.086). Sensitivity analysis demonstrated that no individual SNP significantly influenced the causal inference.

CONCLUSIONS

This study provides evidence of a causal relationship between AF and an elevated risk of stroke. These findings emphasize the significance of managing AF in order to prevent and treat strokes. Additional research is required to better understand the underlying mechanisms of this causal association.

Causal association between constipation and risk of colorectal cancer: a bidirectional two-sample Mendelian randomization study

Background

Colorectal cancer (CRC) is a globally significant health concern, necessitating effective preventive strategies through identifying modifiable risk factors. Constipation, characterized by infrequent bowel movements or difficulty passing stools, has been proposed as a potential CRC risk factor. However, establishing causal links between constipation and CRC remains challenging due to observational study limitations.

Methods

Mendelian randomization (MR) utilizes genetic variants as instrumental variables, capitalizing on genetically determined variation to assess causal relationships. In this dual-sample bidirectional MR study, we extracted genetic data from independent cohorts with CRC (Include colon cancer and rectal cancer) and constipation cases. Genome-wide association studies (GWAS) identified constipation and CRC-associated genetic variants used as instruments to infer causality. The bidirectional MR analysis evaluated constipation's impact on CRC risk and the possibility of reverse causation.

Results

Employing bidirectional MR, we explored the causal relationship between constipation and CRC using publicly available GWAS data. Analysis of constipation's effect on CRC identified 26 significant SNPs, all with strong instrumental validity. IVW-random effect analysis suggested a potential causal link [OR = 1.002(1.000, 1.004); P = 0.023], although alternative MR approaches were inconclusive. Investigating CRC's impact on constipation, 28 significant SNPs were identified, yet IVW analyses found no causal effect [OR = 0.137(0.007, 2.824); P = 0.198]. Other MR methods also yielded no significant causal association. We analyzed constipation separately from colon and rectal cancer using the same methodology in both directions, and no causal relationship was obtained.

Conclusion

Our bidirectional MR study suggests a potential constipation-CRC link, with mixed MR approach outcomes. Limited evidence supports constipation causing CRC. Reliable instruments, minimal heterogeneity, and robust analyses bolster these findings, enriching understanding. Future research should explore additional factors to enhance comprehension and clinical implications.

Uncovering the ceRNA network and DNA methylation associated with gene expression in nasopharyngeal carcinoma

OBJECTIVE

This study aimed to uncover abnormally expressed genes regulated by competitive endogenous RNA (ceRNA) and DNA methylation nasopharyngeal carcinoma and to validate the role of lncRNAs in the ceRNA network on nasopharyngeal carcinoma progression.

METHODS

Based on the GSE64634 (mRNA), GSE32960 (miRNA), GSE95166 (lncRNA), and GSE126683 (lncRNA) datasets, we screened differentially expressed mRNAs, miRNAs and lncRNAs in nasopharyngeal carcinoma. A ceRNA network was subsequently constructed. Differentially methylated genes were screened using the GSE62336 dataset. The abnormally expressed genes regulated by both the ceRNA network and DNA methylation were identified. In the ceRNA network, the expression of RP11-545G3.1 lncRNA was validated in nasopharyngeal carcinoma tissues and cells by RT-qPCR. After a knockdown of RP11-545G3.1, the viability, migration, and invasion of CNE-2 and NP69 cells was assessed by CCK-8, wound healing and Transwell assays.

RESULTS

This study identified abnormally expressed mRNAs, miRNAs and lncRNAs in nasopharyngeal carcinoma tissues. A ceRNA network was constructed, which contained three lncRNAs, 15 miRNAs and 129 mRNAs. Among the nodes in the PPI network based on the mRNAs in the ceRNA network, HMGA1 was assessed in relation to the overall and disease-free survival of nasopharyngeal carcinoma. We screened two up-regulated genes regulated by the ceRNA network and hypomethylation and 26 down-regulated genes regulated by the ceRNA network and hypermethylation. RP11-545G3.1 was highly expressed in the nasopharyngeal carcinoma tissues and cells. Moreover, the knockdown of RP11-545G3.1 reduced the viability, migration, and invasion of CNE-2 and NP69 cells.

CONCLUSION

Our findings uncovered the epigenetic regulation in nasopharyngeal carcinoma and identified the implications of RP11-545G3.1 on the progression of nasopharyngeal carcinoma.

BMI and plasma lipid levels with risk of proliferative diabetic retinopathy: a univariable and multivariable Mendelian randomization study

Background

The study aimed to determine whether a causal effect exists between body mass index (BMI) or plasma lipid levels and proliferative diabetic retinopathy (PDR) risk in humans.

Methods

We utilized univariable (UVMR) and multivariable two-sample Mendelian randomization (MVMR) analyses to confirm the effects of BMI and plasma lipid levels on the risk of PDR. Genetic variants associated with BMI and three plasma lipids were obtained from GWAS summary datasets generated by many different consortia and were deposited in the MR-Base database. The GWAS summary data for PDR from the FinnGen biobank included 2,12,889 participants of European ancestry (8,681 cases and 2,04,208 controls). Inverse variance weighted (IVW) was applied as the main MR analysis. Sensitivity analysis was used to evaluate the robustness of our findings.

Results

In the UVMR analysis, the causal associations of genetically predicted BMI with PDR presented a positive association (OR = 1.120, 95% CI = 1.076-1.167, P < 0.001), and the lower HDL-C level was associated with a higher risk of PDR (OR = 0.898, 95% CI = 0.811-0.995, P = 0.040). No evidence of an association between LDL-C or TG levels (P > 0.05) and PDR risk was found. In the MVMR analysis controlling for the HDL-C level, there was strong evidence for a direct causal effect of BMI on the risk of PDR (OR = 1.106, 95%CI = 1.049, 1.166, P < 0.001, IVW). After adjusting for BMI, there was no evidence for a direct causal effect of the HDL-C level on the risk of PDR (OR = 0.911, 95% CI = 0.823, 1.008, P = 0.072). Sensitivity analyses confirmed that the results were reliable and stable.

Conclusion

Robust evidence was demonstrated for an independent, causal effect of BMI in increasing the risk of PDR. Further studies are required to understand the potential biological mechanisms underlying this causal relationship.

A review of machine learning models applied to genomic prediction in animal breeding

The advent of modern genotyping technologies has revolutionized genomic selection in animal breeding. Large marker datasets have shown several drawbacks for traditional genomic prediction methods in terms of flexibility, accuracy, and computational power. Recently, the application of machine learning models in animal breeding has gained a lot of interest due to their tremendous flexibility and their ability to capture patterns in large noisy datasets. Here, we present a general overview of a handful of machine learning algorithms and their application in genomic prediction to provide a meta-picture of their performance in genomic estimated breeding values estimation, genotype imputation, and feature selection. Finally, we discuss a potential adoption of machine learning models in genomic prediction in developing countries. The results of the reviewed studies showed that machine learning models have indeed performed well in fitting large noisy data sets and modeling minor nonadditive effects in some of the studies. However, sometimes conventional methods outperformed machine learning models, which confirms that there's no universal method for genomic prediction. In summary, machine learning models have great potential for extracting patterns from single nucleotide polymorphism datasets. Nonetheless, the level of their adoption in animal breeding is still low due to data limitations, complex genetic interactions, a lack of standardization and reproducibility, and the lack of interpretability of machine learning models when trained with biological data. Consequently, there is no remarkable outperformance of machine learning methods compared to traditional methods in genomic prediction. Therefore, more research should be conducted to discover new insights that could enhance livestock breeding programs.

Genome-wide association and genomic prediction for a reproductive index summarizing fertility outcomes in U.S. Holsteins

Subfertility represents one major challenge to enhancing dairy production and efficiency. Herein, we use a reproductive index (RI) expressing the predicted probability of pregnancy following artificial insemination (AI) with Illumina 778K genotypes to perform single and multi-locus genome-wide association analyses (GWAA) on 2,448 geographically diverse U.S. Holstein cows and produce genomic heritability estimates. Moreover, we use genomic best linear unbiased prediction (GBLUP) to investigate the potential utility of the RI by performing genomic predictions with cross validation. Notably, genomic heritability estimates for the U.S. Holstein RI were moderate (h2 = 0.1654 ± 0.0317-0.2550 ± 0.0348), while single and multi-locus GWAA revealed overlapping quantitative trait loci (QTL) on BTA6 and BTA29, including the known QTL for the daughter pregnancy rate (DPR) and cow conception rate (CCR). Multi-locus GWAA revealed seven additional QTL, including one on BTA7 (60 Mb) which is adjacent to a known heifer conception rate (HCR) QTL (59 Mb). Positional candidate genes for the detected QTL included male and female fertility loci (i.e. spermatogenesis and oogenesis), meiotic and mitotic regulators, and genes associated with immune response, milk yield, enhanced pregnancy rates, and the reproductive longevity pathway. Based on the proportion of the phenotypic variance explained (PVE), all detected QTL (n = 13; P ≤ 5e - 05) were estimated to have moderate (1.0% < PVE ≤ 2.0%) or small effects (PVE ≤ 1.0%) on the predicted probability of pregnancy. Genomic prediction using GBLUP with cross validation (k = 3) produced mean predictive abilities (0.1692-0.2301) and mean genomic prediction accuracies (0.4119-0.4557) that were similar to bovine health and production traits previously investigated.

Genomic Prediction Methods Accounting for Nonadditive Genetic Effects

The use of genomic information for prediction of future phenotypes or breeding values for the candidates to selection has become a standard over the last decade. However, most procedures for genomic prediction only consider the additive (or substitution) effects associated with polymorphic markers. Nevertheless, the implementation of models that consider nonadditive genetic variation may be interesting because they (1) may increase the ability of prediction, (2) can be used to define mate allocation procedures in plant and animal breeding schemes, and (3) can be used to benefit from nonadditive genetic variation in crossbreeding or purebred breeding schemes. This study reviews the available methods for incorporating nonadditive effects into genomic prediction procedures and their potential applications in predicting future phenotypic performance, mate allocation, and crossbred and purebred selection. Finally, a brief outline of some future research lines is also proposed.

The Causal Relationship Between Body Mass Index and the Risk of Osteoarthritis

Objective

The study aimed to explore the causal effect of body mass index (BMI) on osteoarthritis.

Methods

The genome-wide association data of BMI and osteoarthritis were obtained via the Mendelian randomization (MR)-base platform. Single nucleotide polymorphisms (SNPs) significantly associated with BMI were identified and used as instrumental variables, and the causal relationship between BMI and osteoarthritis was examined using the two-sample MR research method. Three statistical methods including inverse-variance weighted (IVW) method, weighted median estimator, and MR-Egger regression were employed.

Results

A total of 79 SNPs significantly associated with BMI were identified in the study (P<5×10⁻⁸; linkage disequilibrium r² <0.1). Consistent association between BMI and osteoarthritis was observed when evaluated by different methods (IVW: odds ratio (OR) 1.028, 95% confidence interval (CI) 1.021–1.036; weighted median estimator: OR 1.028, 95% CI 1.019–1.037; MR-Egger regression: OR 1.028, 95% CI 1.009–1.046), which suggests that BMI is positively associated with increased risk of osteoarthritis. There was no evidence that the observed causal effect between BMI and the risk of osteoarthritis was affected by genetic pleiotropy (MR-Egger intercept 1.3×10⁻⁵, P=0.959).

Conclusion

The MR analysis provided the strong evidence to indicate that BMI might be causally associated with the risk of osteoarthritis.

Cross-validation of best linear unbiased predictions of breeding values using an efficient leave-one-out strategy

Empirical estimates of the accuracy of estimates of breeding values (EBV) can be obtained by cross-validation. Leave-one-out cross-validation (LOOCV) is an extreme case of k-fold cross-validation. Efficient strategies for LOOCV of predictions of phenotypes have been developed for a simple model with an overall mean and random marker or animal genetic effects. The objective here was to develop and evaluate an efficient LOOCV method for prediction of breeding values and other random effects under a general mixed linear model with multiple random effects. Conventional LOOCV of EBV requires inverting an (n-1)×(n-1) covariance matrix for each of n (= number of observations) data sets. Our efficient LOOCV obtains the required inverses from the inverse of the covariance matrix for all n observations. The efficient method can be applied to complex models with multiple fixed and random effects, but requires fixed effects to be treated as random, with large variances. An alternative is to precorrect observations using estimates of fixed effects obtained from the complete data, but this can lead to biases. The efficient LOOCV method was compared to conventional LOOCV of predictions of breeding values in terms of computational demands and accuracy. For a data set with 3,205 observations and a model with multiple random and fixed effects, the efficient LOOCV method was 962 times faster than the conventional LOOCV with precorrection for fixed effects based on each training data set but resulted in identical EBV. A computationally efficient LOOCV for prediction of breeding values for single- and multiple-trait mixed models with multiple fixed and random effects was successfully developed. The method enables cross-validation of predictions of breeding values and of any linear combination of random and/or fixed effects, along with leave-one-out precorrection of validation phenotypes.

A review of traditional and machine learning methods applied to animal breeding

The current livestock management landscape is transitioning to a high-throughput digital era where large amounts of information captured by systems of electro-optical, acoustical, mechanical, and biosensors is stored and analyzed on a daily and hourly basis, and actionable decisions are made based on quantitative and qualitative analytic results. While traditional animal breeding prediction methods have been used with great success until recently, the deluge of information starts to create a computational and storage bottleneck that could lead to negative long-term impacts on herd management strategies if not handled properly. A plethora of machine learning approaches, successfully used in various industrial and scientific applications, made their way in the mainstream approaches for livestock breeding techniques, and current results show that such methods have the potential to match or surpass the traditional approaches, while most of the time they are more scalable from a computational and storage perspective. This article provides a succinct view on what traditional and novel prediction methods are currently used in the livestock breeding field, how successful they are, and how the future of the field looks in the new digital agriculture era.

Waist circumference increases risk of coronary heart disease: Evidence from a Mendelian randomization study

BACKGROUND

This study investigated whether expanding waist circumference (WC) is causally associated with an elevated risk of coronary heart disease (CHD), using a two-sample Mendelian randomization (MR) study through integrating summarized data from genome-wide association study.

METHODS

The data included in this analysis were mainly from the Genetic Investigation of ANthropometric Traits (GIANT), Consortium and Coronary Artery Disease Genome wide Replication, and Meta-analysis plus the Coronary Artery Disease (C4D) Genetics (CARDIoGRAMplusC4D) Consortium. Three statistical approaches, inverse-variance weighted (IVW), weighted median, and MR-Egger regression method were conducted to assess the casual relationship. The exposure was WC, measured by 46 single-nucleotide polymorphisms from GIANT and the outcome was the risk of CHD. Then, we used the genetic data from Neale Lab and TAG to infer whether WC causally affected the established risk factors of CHD.

RESULTS

The IVW method presented that genetically predicted WC was positively casually associated with CHD (odds ratio [OR]: 1.57, 95% CI = 1.33-1.84; p = 4.81e-08), which was consistent with the result of weighted median and MR-Egger regression. MR-Egger regression indicated that there was no directional horizontal pleiotropy to violate the MR assumption. Additionally, expanded WC was also associated with higher risk of hypertension and diabetes, higher cholesterol, more smoking intensity, and decreased frequency of physical activity.

CONCLUSION

Our analysis provided strong evidence to indicate a causal relationship between WC and increased risk of CHD.

Behavior of the Linear Regression method to estimate bias and accuracies with correct and incorrect genetic evaluation models

Bias in genetic evaluations has been a constant concern in animal genetics. The interest in this topic has increased in the last years, since many studies have detected overestimation (bias) in estimated breeding values (EBV). Detecting the existence of bias, and the realized accuracy of predictions, is therefore of importance, yet this is difficult when studying small data sets or breeds. In this study, we tested by simulation the recently presented method Linear Regression (LR) for estimation of bias, slope, and accuracy of pedigree EBV. The LR method computes statistics by comparing EBV from a data set containing old, partial information with EBV from a data set containing all information (old and new, a whole data set) for the same individuals. The method proposes an estimator for bias (Δ_pˆ), an estimator of slope (b_pˆ), and 3 estimators related to accuracies: the ratio between accuracies [Formula: see text] the reliability of the partial data set (acc_p²ˆ), and the ratio of reliabilities (ρ_p,w²ˆ). We simulated a dairy scheme for low (0.10) and moderate (0.30) heritabilities. In both cases, we checked the behavior of the estimators for 3 scenarios: (1) when the evaluation model is the same as the model used to simulate the data; (2) when the evaluation model uses an incorrect heritability; and (3) when the data includes an environmental trend. For scenarios in which the evaluation model was correct, the LR method was capable of correctly estimating bias, slope, and accuracies, with better performance for higher heritability [i.e., corr(b_p,b_pˆ) was 0.45 for h² = 0.10 and 0.59 for h² = 0.30]. In cases of the use of incorrect heritabilities in the evaluation model, the bias was correctly estimated in direction but not in magnitude. In the same way, the magnitudes of bias and of slope were underestimated in scenarios with environmental trends in data, except for cases in which contemporary groups were random and greatly shrunken. In general, accuracies were well estimated in all scenarios. The LR method is capable of checking bias and accuracy in all cases, if the evaluation model is reasonably correct or robust, and its estimations are more precise with more information (e.g., high heritability). If the model uses an incorrect heritability or a hidden trend exists in the data, it is still possible to estimate the direction and existence of bias and slope but not always their magnitudes.

Genomic merit for reproductive traits. II: Physiological responses of Holstein heifers

Fertility traits were recently added to the evaluation of genetic merit, allowing for the selection of Holstein cattle with improved reproductive performance. In the current study, we investigated the associations among genomic merit for daughter pregnancy rate (GDPR) and heifer conception rate (GHCR) and physiological responses during proestrus and diestrus. Holstein heifers (n = 99) were classified based on GDPR [high = 3.26 ± 0.76 (1.6 to 5.3), n = 48; low = -0.17 ± 0.75 (-1.8 to 1.0), n = 51] and GHCR [high = 2.75 ± 0.77 (1.5 to 5.5), n = 49; low = 0.06 ± 0.67 (-2.1 to 1.2), n = 50]. Heifers were fitted with an automated estrous detection device, were treated with PGF_2α for synchronization of estrus, and received either artificial insemination or embryo transfer at detected estrus. Blood was sampled at the time of PGF_2α treatment, within 24 h of the onset of estrus (d 0), and on d 7, 14, 19 ± 2, 28, and 35. Blood samples from all heifers were analyzed for concentrations of estradiol (d 0) and progesterone (on the day of PGF_2α treatment and d 0, 7, and 14). Blood samples from heifers pregnant on d 38 ± 3 were analyzed for concentrations of progesterone (d 0, 7, 14, 19 ± 2, 28, and 35), pregnancy-specific protein B (d 19 ± 2, 28, and 35), and insulin-like growth factor 1 (d 0, 7, 14, 19 ± 2, 28, and 35). Expression of mRNA for interferon-stimulated gene 15 in peripheral leukocytes isolated from blood collected on d 19 ± 2 was determined. Ovaries were scanned by ultrasound daily from d 0 to 4 or until ovulation was detected. Heifers with low GHCR tended to be less likely to be detected in estrus (78.0 vs. 91.8%). Estradiol concentration on d 0 was greater for heifers with high GDPR (4.53 ± 0.23 vs. 3.79 ± 0.23 pg/mL). The ovulatory follicle was larger for heifers with high GDPR (16.28 ± 0.33 vs. 14.55 ± 0.35 mm), whereas heifers with high GHCR tended to have smaller ovulatory follicles (15.00 ± 0.31 vs. 15.83 ± 0.37 mm). Heifers with high GDPR tended to be more likely to ovulate within 96 h of the onset of estrus (90.7 vs. 75.0%). Among heifers pregnant on d 38 ± 3, GDPR and GHCR were not associated with mRNA expression for interferon-stimulated gene 15. Heifers with high GDPR had greater concentration of pregnancy-specific protein B from d 28 to 35 (3.03 ± 0.15 vs. 2.48 ± 0.1 ng/mL). Heifers with high GHCR tended to have greater insulin-like growth factor 1 concentration from d 7 to 35 (108.0 ± 3.2 vs. 97.7 ± 4.2 ng/mL). Our results suggest that selection for Holstein cattle for GDPR may have positive effects on reproductive performance through changes in ovarian follicle development and steroidogenesis. Although selection of Holstein cattle for GHCR may negatively affect estrous expression by affecting ovarian follicle growth, selection for GHCR may improve reproductive performance by altering the somatotropic axis.

A random forest classifier predicts recurrence risk in patients with ovarian cancer

Ovarian cancer (OC) is associated with a poor prognosis due to difficulties in early detection. The aims of the present study were to construct a recurrence risk prediction model and to reveal important OC genes or pathways. RNA sequencing data was obtained for 307 OC samples, and the corresponding clinical data were downloaded from The Cancer Genome Atlas database. Additionally, two validation datasets, GSE44104 (20 recurrent and 40 non-recurrent OC samples) and GSE49997 (204 OC samples), were obtained from the Gene Expression Omnibus database. Differentially expressed genes were screened using the differential expression via distance synthesis algorithm, followed by gene ontology enrichment analysis and weighted gene coexpression network analysis (WGCNA). Furthermore, subnetwork analysis was conducted for the protein-protein interaction (PPI) network using the BioNet package. Finally, a random forest classifier was constructed based on the subnetwork nodes, and its reliability was validated using the GSE44104 and GSE49997 validation datasets. A total of 44 upregulated and 117 downregulated genes were identified in the recurrent samples. Enrichment analysis indicated that cytochrome P450 family 17 subfamily A member 1 (CYP17A1) was associated with ‘positive regulation of steroid hormone biosynthetic processes’. WGCNA identified turquoise and grey modules that were significantly correlated with status and prognosis. A significant PPI subnetwork containing 16 nodes was also identified, including: Transcription factor GATA-4; fibroblast growth factor 9; aromatase; 3β-hydroxysteroid dehydrogenase/δ5-4-isomerase type 2; corticosteroid 11β-dehydrogenase isozyme 1; CYP17A1; pituitary homeobox 2; left-right determination factor 1; homeobox protein ARX; estrogen receptor β; steroidogenic factor 1; forkhead box protein L2; myocardin; steroidogenic acute regulatory protein mitochondrial; vesicular inhibitory amino acid transporter; and twist-related protein 1. A random forest classifier was constructed using the subnetwork nodes as feature genes, which exhibited a 92% true positive rate when classifying recurrent and non-recurrent OC samples. The classifying efficiency of the random forest classifier was validated using the two other independent datasets. Overall, 44 upregulated and 117 downregulated genes associated with OC recurrence were identified. Furthermore, the 16 subnetwork node genes that were identified may be important molecules in OC recurrence.

In Vitro Micropatterned Human Pluripotent Stem Cell Test (µP-hPST) for Morphometric-Based Teratogen Screening

Exposure to teratogenic chemicals during pregnancy may cause severe birth defects. Due to high inter-species variation of drug responses as well as financial and ethical burdens, despite the widely use of in vivo animal tests, it’s crucial to develop highly predictive human pluripotent stem cell (hPSC)-based in vitro assays to identify potential teratogens. Previously we have shown that the morphological disruption of mesoendoderm patterns formed by geometrically-confined cell differentiation and migration using hPSCs could potentially serve as a sensitive morphological marker in teratogen detection. Here, a micropatterned human pluripotent stem cell test (µP-hPST) assay was developed using 30 pharmaceutical compounds. A simplified morphometric readout was developed to quantify the mesoendoderm pattern changes and a two-step classification rule was generated to identify teratogens. The optimized µP-hPST could classify the 30 compounds with 97% accuracy, 100% specificity and 93% sensitivity. Compared with metabolic biomarker-based hPSC assay by Stemina, the µP-hPST could successfully identify misclassified drugs Bosentan, Diphenylhydantoin and Lovastatin, and show a higher accuracy and sensitivity. This scalable µP-hPST may serve as either an independent assay or a complement assay for existing assays to reduce animal use, accelerate early discovery-phase drug screening and help general chemical screening of human teratogens.