Arjun Krishnan

A module-based approach for post-omics, post-GWAS network-based gene classification

The paper introduces ModGenePlexus, a network-based gene classification method that improves disease gene discovery by decomposing gene lists into modules and combining module-specific classifiers.

Large-scale automated detection reveals pervasive sex imbalance in biomedical research

This paper quantifies sex imbalance in biomedical research by analyzing transcriptome samples and study metadata, revealing male-skewed representation in many disease areas.

Language Model Embedding Classifiers Enable Identification of Multiple Sclerosis-Associated BCRs and Repertoires

The paper develops a language model to identify MS-associated BCRs by embedding BCR sequences and classifying them for disease association.

Linking Genetic Risk to Disease-Relevant Cellular States via Metacell-Informed Modeling with ICePop

ICePop links genetic risk to disease-relevant cellular states by analyzing single-cell transcriptomics data across multiple traits, identifying 1,684 disease-cell type associations.

Examining the Genetic Links between Clusters of Immune-mediated Diseases and Psychiatric Disorders

This study identifies genetic links between clusters of immune-mediated diseases and psychiatric disorders using genomic structural equation modeling and a new heterogeneity metric.

Integrative transcriptome-based drug repurposing in tuberculosis

This study identifies 64 FDA-approved drugs as potential host-directed therapeutics for tuberculosis using an integrative transcriptome-based approach combining multiple gene expression signatures and connectivity scoring methods.

Annotating publicly-available samples and studies using interpretable modeling of unstructured metadata

This paper introduces txt2onto 2.0, a method for annotating biomedical metadata with disease and tissue terms using interpretable machine learning, showing improved performance over its predecessor.

Uncovering hidden gene-trait patterns through biclustering analysis of the UK Biobank

The paper applies biclustering to GWAS data from the UK Biobank to uncover local gene-trait associations that traditional methods miss.

Network-based representation learning reveals the impact of age and diet on the gut microbial and metabolomic environment of U.S. infants in a randomized controlled feeding trial

The paper uses network representation learning to analyze gut microbiome and metabolomic data from U.S. infants, finding diet type influences microbial and metabolic profiles during complementary feeding.

Joint representation of molecular networks from multiple species improves gene classification

GenePlexusZoo improves gene classification by integrating molecular networks from multiple species into a shared feature space, enhancing both single-species and cross-species predictions.

CONE: COntext-specific Network Embedding via Contextualized Graph Attention

CONE generates context-specific gene embeddings from context-free interactomes to identify disease-associated genes using contextualized random walks and graph attention networks.

PyGenePlexus: A Python package for gene discovery using network-based machine learning

PyGenePlexus predicts gene associations with input gene sets using network-based machine learning and provides interpretability through comparison with known gene sets.

Human pan-body age- and sex-specific molecular phenomena inferred from public transcriptome data using machine learning

The paper uses machine learning on public transcriptome data to predict age and sex from gene expression and identify molecular signatures associated with age and sex differences.

Open Biomedical Network Benchmark: A Python Toolkit for Benchmarking Datasets with Biomedical Networks

The paper introduces OBNB, a benchmarking resource for GNN-based network biology tasks, providing datasets and a Python toolkit for evaluating gene classification models.

Accurately modeling biased random walks on weighted networks using node2vec+

The paper introduces node2vec+, an improved version of node2vec that incorporates edge weights in biased random walks, showing better performance on weighted gene networks.

A network-based approach for isolating the chronic inflammation gene signatures underlying complex diseases towards finding new treatment opportunities

The paper presents a computational approach to identify gene signatures of chronic inflammation in complex diseases and prioritize drugs for new therapeutic opportunities.

Robust normalization and transformation techniques for constructing gene coexpression networks from RNA-seq data

This study benchmarks 30 RNA-seq data processing workflows for constructing coexpression networks, finding that normalization methods significantly affect network accuracy.

Systematic tissue annotations of -omics samples by modeling unstructured metadata

The paper develops an NLP-ML approach to annotate -omics samples with tissue and cell-type labels using free-text metadata, outperforming existing methods and demonstrating biological relevance.

Functional assessment of the "two-hit" model for neurodevelopmental defects in Drosophila and X. laevis

The paper tests a 'two-hit' model for neurodevelopmental defects by examining interactions between 16p12.1 homologs and 'second-hit' genes in Drosophila and Xenopus laevis, finding genetic interactions in 63% of tested pairs.

PecanPy: a fast, efficient, and parallelized Python implementation of node2vec

PecanPy provides a fast and efficient Python implementation of node2vec for embedding nodes in biological networks, improving scalability for dense graphs.

Combinatorial patterns of gene expression changes contribute to variable expressivity of the developmental delay-associated 16p12.1 deletion

The paper investigates how gene expression changes from a 16p12.1 deletion and second-hit variants contribute to phenotypic variability in developmental delay.

A Flexible, Interpretable, and Accurate Approach for Imputing the Expression of Unmeasured Genes

This paper introduces SampleLASSO, a method for imputing unmeasured gene expression in microarray data, showing it outperforms existing methods across multiple tasks.

Co-expression signatures of combinatorial gene regulation

The paper shows that TF-TF interactions improve co-expression signals for predicting regulatory complexes in Arabidopsis, highlighting the role of combinatorial gene regulation.

Supervised-learning is an accurate method for network-based gene classification

This study benchmarks supervised-learning for network-based gene classification, showing it outperforms label-propagation and node-embeddings in accuracy.

Rare variants in the genetic background modulate the expressivity of neurodevelopmental disorders

The paper shows that rare secondary mutations in functionally intolerant genes affect the expressivity of neurodevelopmental disorders in individuals with primary mutations.

SANe: The Seed Active Network for Discovering Transcriptional Regulatory Programs of Seed Development.

The paper develops a genome-scale regulatory network for Arabidopsis seed development, identifies key transcription factors, and provides a webtool for analyzing transcriptomes and predicting TF regulatory effects.

Pervasive epistasis modulates neurodevelopmental defects of the autism-associated 16p11.2 deletion

The paper identifies genetic interactions within the 16p11.2 deletion that modulate neurodevelopmental phenotypes in Drosophila and links these interactions to human brain networks.

RECoN: Rice Environment Coexpression Network for Systems Level Analysis of Abiotic-Stress Response

RECoN is a web resource that uses gene coexpression networks to analyze abiotic stress responses in rice, revealing functional and regulatory insights from new transcriptome datasets.