An expression-directed linear mixed model discovering low-effect genetic variants.

Detecting genetic variants with low-effect sizes using a moderate sample size is difficult, hindering downstream efforts to learn pathology and estimating heritability. In this work, by utilizing informative weights learned from training genetically predicted gene expression models, we formed an alternative approach to estimate the polygenic term in a linear mixed model. Our linear mixed model estimates the genetic background by incorporating their relevance to gene expression. Our protocol, expression-directed linear mixed model, enables the discovery of subtle signals of low-effect variants using moderate sample size. By applying expression-directed linear mixed model to cohorts of around 5,000 individuals with either binary (WTCCC) or quantitative (NFBC1966) traits, we demonstrated its power gain at the low-effect end of the genetic etiology spectrum. In aggregate, the additional low-effect variants detected by expression-directed linear mixed model substantially improved estimation of missing heritability. Expression-directed linear mixed model moves precision medicine forward by accurately detecting the contribution of low-effect genetic variants to human diseases.

cLD: Rare-variant linkage disequilibrium between genomic regions identifies novel genomic interactions.

Linkage disequilibrium (LD) is a fundamental concept in genetics; critical for studying genetic associations and molecular evolution. However, LD measurements are only reliable for common genetic variants, leaving low-frequency variants unanalyzed. In this work, we introduce cumulative LD (cLD), a stable statistic that captures the rare-variant LD between genetic regions, which reflects more biological interactions between variants, in addition to lack of recombination. We derived the theoretical variance of cLD using delta methods to demonstrate its higher stability than LD for rare variants. This property is also verified by bootstrapped simulations using real data. In application, we find cLD reveals an increased genetic association between genes in 3D chromatin interactions, a phenomenon recently reported negatively by calculating standard LD between common variants. Additionally, we show that cLD is higher between gene pairs reported in interaction databases, identifies unreported protein-protein interactions, and reveals interacting genes distinguishing case/control samples in association studies.

XA4C: eXplainable representation learning via Autoencoders revealing Critical genes.

Machine Learning models have been frequently used in transcriptome analyses. Particularly, Representation Learning (RL), e.g., autoencoders, are effective in learning critical representations in noisy data. However, learned representations, e.g., the "latent variables" in an autoencoder, are difficult to interpret, not to mention prioritizing essential genes for functional follow-up. In contrast, in traditional analyses, one may identify important genes such as Differentially Expressed (DiffEx), Differentially Co-Expressed (DiffCoEx), and Hub genes. Intuitively, the complex gene-gene interactions may be beyond the capture of marginal effects (DiffEx) or correlations (DiffCoEx and Hub), indicating the need of powerful RL models. However, the lack of interpretability and individual target genes is an obstacle for RL's broad use in practice. To facilitate interpretable analysis and gene-identification using RL, we propose "Critical genes", defined as genes that contribute highly to learned representations (e.g., latent variables in an autoencoder). As a proof-of-concept, supported by eXplainable Artificial Intelligence (XAI), we implemented eXplainable Autoencoder for Critical genes (XA4C) that quantifies each gene's contribution to latent variables, based on which Critical genes are prioritized. Applying XA4C to gene expression data in six cancers showed that Critical genes capture essential pathways underlying cancers. Remarkably, Critical genes has little overlap with Hub or DiffEx genes, however, has a higher enrichment in a comprehensive disease gene database (DisGeNET) and a cancer-specific database (COSMIC), evidencing its potential to disclose massive unknown biology. As an example, we discovered five Critical genes sitting in the center of Lysine degradation (hsa00310) pathway, displaying distinct interaction patterns in tumor and normal tissues. In conclusion, XA4C facilitates explainable analysis using RL and Critical genes discovered by explainable RL empowers the study of complex interactions.

Enterolactone and trabectedin suppress epithelial ovarian cancer synergistically via upregulating THBS1.

Epithelial ovarian cancer (EOC) is the most common and fatal subtype of ovarian malignancies, with no effective therapeutics available. Our previous studies have demonstrated extraordinary suppressive efficacy of enterolactone (ENL) on EOC. A chemotherapeutic agent, trabectedin (Trabe), is shown to be effective on ovarian cancer, especially when combined with other therapeutics, such as pegylated liposomal doxorubicin or oxaliplatin. Thrombospondin 1 (THBS1), a kind of matrix glycoprotein, plays important roles against cancer development through inhibiting angiogenesis but whether it is involved in the suppression of EOC by ENL or Trabe remains unknown. To test combined suppressive effects of ENL and Trabe on EOC and possible involvement of THBS1 in the anticancer activities of ENL and Trabe. The EOC cell line ES-2 was transfected with overexpressed THBS1 by lentivirus vector. We employed tube formation assay to evaluate the anti-angiogenesis activity of ENL and of its combined use with Trabe after THBS1 overexpression and established drug intervention and xenograft nude mouse cancer models to assess the in vivo effects of the hypothesized synergistic suppression between the agents and the involvement of THBS1. Mouse fecal samples were collected for 16S rDNA sequencing and microbiota analysis. We detected strong inhibitory activities of ENL and Trabe against the proliferation and migration of cancer cells and observed synergistic effects between ENL and Trabe in suppressing EOC. ENL and Trabe, given either separately or in combination, could suppress the tube formation capability of human microvascular endothelial cells, and this inhibitory effect became even stronger with THBS1 overexpression. In the ENL plus Trabe combination group, the expression of tissue inhibitor of metalloproteinases 3 and cluster of differentiation 36 was both upregulated, whereas matrix metalloproteinase 9, vascular endothelial growth factor, and cluster of differentiation 47 were all decreased. With the overexpression of THBS1, the results became even more pronounced. In animal experiments, combined use of ENL and Trabe showed superior inhibitory effects to either single agent and significantly suppressed tumor growth, and the overexpression of THBS1 further enhanced the anti-cancer activities of the drug combination group. ENL and Trabe synergistically suppress EOC and THBS1 could remarkably facilitate the synergistic anticancer effects of ENL and Trabe.

Stabilized COre gene and Pathway Election uncovers pan-cancer shared pathways and a cancer-specific driver.

Approaches systematically characterizing interactions via transcriptomic data usually follow two systems: (i) coexpression network analyses focusing on correlations between genes and (ii) linear regressions (usually regularized) to select multiple genes jointly. Both suffer from the problem of stability: A slight change of parameterization or dataset could lead to marked alterations of outcomes. Here, we propose Stabilized COre gene and Pathway Election (SCOPE), a tool integrating bootstrapped least absolute shrinkage and selection operator and coexpression analysis, leading to robust outcomes insensitive to variations in data. By applying SCOPE to six cancer expression datasets (BRCA, COAD, KIRC, LUAD, PRAD, and THCA) in The Cancer Genome Atlas, we identified core genes capturing interaction effects in crucial pan-cancer pathways related to genome instability and DNA damage response. Moreover, we highlighted the pivotal role of CD63 as an oncogenic driver and a potential therapeutic target in kidney cancer. SCOPE enables stabilized investigations toward complex interactions using transcriptome data.

Disentangling genetic feature selection and aggregation in transcriptome-wide association studies.

The success of transcriptome-wide association studies (TWAS) has led to substantial research toward improving the predictive accuracy of its core component of genetically regulated expression (GReX). GReX links expression information with genotype and phenotype by playing two roles simultaneously: it acts as both the outcome of the genotype-based predictive models (for predicting expressions) and the linear combination of genotypes (as the predicted expressions) for association tests. From the perspective of machine learning (considering SNPs as features), these are actually two separable steps-feature selection and feature aggregation-which can be independently conducted. In this study, we show that the single approach of GReX limits the adaptability of TWAS methodology and practice. By conducting simulations and real data analysis, we demonstrate that disentangled protocols adapting straightforward approaches for feature selection (e.g., simple marker test) and aggregation (e.g., kernel machines) outperform the standard TWAS protocols that rely on GReX. Our development provides more powerful novel tools for conducting TWAS. More importantly, our characterization of the exact nature of TWAS suggests that, instead of questionably binding two distinct steps into the same statistical form (GReX), methodological research focusing on optimal combinations of feature selection and aggregation approaches will bring higher power to TWAS protocols.

kTWAS: integrating kernel machine with transcriptome-wide association studies improves statistical power and reveals novel genes.

The power of genotype-phenotype association mapping studies increases greatly when contributions from multiple variants in a focal region are meaningfully aggregated. Currently, there are two popular categories of variant aggregation methods. Transcriptome-wide association studies (TWAS) represent a set of emerging methods that select variants based on their effect on gene expressions, providing pretrained linear combinations of variants for downstream association mapping. In contrast to this, kernel methods such as sequence kernel association test (SKAT) model genotypic and phenotypic variance use various kernel functions that capture genetic similarity between subjects, allowing nonlinear effects to be included. From the perspective of machine learning, these two methods cover two complementary aspects of feature engineering: feature selection/pruning and feature aggregation. Thus far, no thorough comparison has been made between these categories, and no methods exist which incorporate the advantages of TWAS- and kernel-based methods. In this work, we developed a novel method called kernel-based TWAS (kTWAS) that applies TWAS-like feature selection to a SKAT-like kernel association test, combining the strengths of both approaches. Through extensive simulations, we demonstrate that kTWAS has higher power than TWAS and multiple SKAT-based protocols, and we identify novel disease-associated genes in Wellcome Trust Case Control Consortium genotyping array data and MSSNG (Autism) sequence data. The source code for kTWAS and our simulations are available in our GitHub repository (https://github.com/theLongLab/kTWAS).

Evaluation of A Phylogenetic Pipeline to Examine Transmission Networks in A Canadian HIV Cohort.

Keywords: HIV; Canada; molecular phylogenetics; viral evolution; person-to-person transmission inference; transmission network; summary statistics.