Biology subjects

Jha, I. P.

Publications and source records attributed to Jha, I. P..

4 recordsLinked to original sources

bioRxiv2026

Inferring large networks with matrix factorisation to capture non-linear dependencies among genes using sparse single-cell profiles

Inference of non-linear dependencies among a large number of features from their scores is an unresolved challenge. Especially when the feature-score matrix is very sparse, like single-cell transcriptome profiles, the problem of estimating dependencies among a large number of features to infer a network becomes an even more daunting task. Here, we propose a method of network inference in reduced dimension (NIRD) to handle sparsity and computational complexity while still inferring non-linear dependencies among genes (features) using large, sparse gene-expression matrices (feature scores). Our method is based on matrix factorisation of gene-expression matrix to facilitate internal imputation as well as network inference using tree ensemble-based non-linear regression. NIRD not only outperformed many other methods across multiple single-cell transcriptomic profiles but also provided consistent inferred networks even in the presence of batch effects. The consistency provided by NIRD helps compare inferred networks to identify genuine genes responsible for changes in regulation due to disease or stress. NIRD can also be used with RNA velocity for better inference of non-linear causality. Application of NIRD with RNA-velocity could improve the prediction of direct targets of transcription factors in human embryonic stem cells, which we validated using ChIP-seq and gene-knockout datasets.

bioinformatics↗

bioRxiv2023

Deciphering the phenotypic heterogeneity and drug response in cancer cells using genome-wide activity and interaction of chromatin domains.

The effect of co-localization of genes in the topologically associated domains (TADs) and their activity as a regulatory unit in cancer samples and cells, together with drug-response, needs comprehensive analysis. Here, we analyzed the activity of TADs using cancer-cell transcriptomes along with chromatin-interaction and epigenome profiles to understand their relationship with drug-response. Our analysis of 819 cancer cell-line transcriptomes revealed that their response to multiple drugs was more correlated with the activity of individual TADs than genes. Applying our approach to 9014 cancer patients data (20 different cancer types) also revealed a higher association between survival and the activity of thousands of individual TADs in comparison to their genes. CRISPR-mediated knock-out of regulatory sites inside a TAD associated with cisplatin-response of oral cancer cells and discovery of primate-specific gain of synteny of genes within a TAD containing EGFR gene and its contribution towards cancer malignancy demonstrate greater utility of TAD-activity based analysis.

cancer biology↗

bioRxiv2022

Inferring functions of coding and non-coding genes using epigenomic patterns and deciphering the effect of combinatorics of transcription factors binding at promoters

The number of annotated genes in the human genome has increased tremendously, and understanding their biological role is challenging through experimental methods alone. There is a need for a computational approach to infer the function of genes, particularly for non-coding RNAs, with reliable explainability. We have utilized genomic features that are present across both coding and non-coding genes like transcription factor (TF) binding pattern, histone modifications, and DNase hypersensitivity profiles to predict ontology-based functions of genes. Our approach for gene function prediction (GFPred) made reliable predictions (>90% balanced accuracy) for 486 gene-sets. Further analysis revealed that predictability using only TF-binding patterns at promoters is also high, and it paved the way for studying the effect of their combinatorics. The predicted associations between functions and genes were validated for their reliability using PubMed abstract mining. Clustering functions based on shared top predictive TFs revealed many latent groups of gene-sets involved in common major biological processes. Available CRISPR screens also supported the inferred association of genes with the major biological processes of latent groups of gene-sets. For the explainability of our approach, we also made more insights into the effect of combinatorics of TF binding (especially TF-pairs) on association with biological functions.

bioinformatics↗

bioRxiv2019

Distance preserving dimension reduction withlocal-topology based scaling for improvedclassification of Biomedical data-sets.

Dimension reduction is often used for several procedures of analysis of high dimensional biomedical data-sets such as classification or outlier detection. To improve performance of such data-mining steps, preserving both distance information and local topology among data-points could be more useful than giving priority to visualisation in low dimension. Therefore, we introduce topology preserving distance scaling (TPDS) to augment dimension reduction method meant to reproduce distance information in higher dimension. Our approach involves distance inflation to preserve local topology to avoid collapse during distance preservation based optimisation. Applying TPDS on diverse biomedical data-sets revealed that besides providing better visualisation than typical distance preserving methods, TPDS leads to better classification of data points in reduced dimension. For data-sets with outliers, the approach of TPDS also proves to be useful, even for purely distance-preserving method for achieving better convergence.

bioinformatics↗