Biology subjects

Jha, A.

Publications and source records attributed to Jha, A..

61 records · Page 4Linked to original sources

bioRxiv2019

Improving interpretability of deep learning models: splicing codes as a case study

Despite the success and fast adaptation of deep learning models in a wide range of fields, lack of interpretability remains an issue, especially in biomedical domains. A recent promising method to address this limitation is Integrated Gradients (IG), which identifies features associated with a prediction by traversing a linear path from a baseline to a sample. We extend IG with nonlinear paths, embedding in latent space, alternative baselines, and a framework to identify important features which make it suitable for interpretation of deep models for genomics.

bioinformatics↗

bioRxiv2019

Environmental niche forecasts of globally threatened Yellow-throated Bulbul, Pycnonotus xantholaemus for conservation prospects in the Deccan peninsula, India

Yellow-throated Bulbul, Pycnonotus xantholaemus is an endemic, rare and threatened species with discontinuous distribution across Deccan peninsula. IUCN has enlisted surveys for newer populations as one of the conservation measures. We used maxent algorithm to generate environmental niche model for further surveys. We looked into climatic envelope at occurrence points and contrast it with background climatic envelope. We compared the model for current scenario and future scenario to assess change in extent of predicted niche over time. We used six variables: climatic, topographical and vegetation layers, and a final set of 102 verified presence locations to generate the model. Topographic ruggedness index and precipitation of wettest month (Bio13) were the strong predictors for Yellow-throated Bulbul niche. Model predicts highly discontinuous and small fragments totalling 7% area of peninsular India as suitable niche. Only 10 % of predicted niche falls within Indias protected area network. Loss of habitat due to granite quarrying and anthropogenic pressure will be a bigger threat than climate change.

ecology↗

bioRxiv2018

Bedsect: An integrated web server application to perform intersection, visualization and functional annotation of genomic regions from multiple datasets

A large number of genomic regions captured during next generation sequencing (NGS) data analyses are generally overlapped to answer a variety of biological questions. Though several command-line tools are available to perform such analysis, there is lack of a comprehensive web server application to perform the genomic region intersections from multiple datasets, to generate a plot for the subsets of the overlapped regions and to perform integrated functional annotation. To address this gap, we have developed a user-friendly integrated web server application i.e. Bedsect, where users can upload genomic regions from multiple datasets without any file number limitation and perform intersection analysis along with visualization of the intersection regions as UpSet and Correlation plot using integrated Shiny application. Bedsect also integrates GREAT for functional annotation, gene ontology, and biological pathways enrichment analysis from identified unique as well as intersected genomic regions. These genomic regions can be further uploaded in the UCSC genome browser for visualization of the results as custom tracks directly from the tool. Bedsect is available at http://imgsb.org/bedsect/.

bioinformatics↗

bioRxiv2018

Widespread alteration of protein autoinhibition in human cancers

Autoinhibition is a prevalent allosteric regulatory mechanism in signaling proteins as it prevents spurious pathway activation and primes for signal propagation only under appropriate inputs. Altered functioning of inhibitory allosteric switches underlies the tumorigenic potential of numerous cancer drivers. However, whether protein autoinhibition is altered generically in cancer cells remains elusive. Here, we reveal that cancer-associated missense mutations and fusion breakpoints are found with significant enrichment within inhibitory allosteric switches across all cancer types, which in the case of the fusion breakpoints is specific to cancer and not present in other diseases. Recurrently disrupted or mutated allosteric switches identify established and new cancer drivers. Cancer-specific mutations in allosteric switches are associated with distinct changes in signaling, and suggest molecular mechanisms for altered protein regulation, which in the case of ASK1, DAPK2 and EIF4G1 were supported by biophysical simulations. Our results demonstrate that autoinhibition-modulating genetic alterations are positively selected for by cancer cells, and that their study provides valuable insights into molecular mechanisms of cancer misregulation.

cancer biology↗

bioRxiv2018

Q-FADD: A mechanistic approach for modeling the accumulation of proteins at sites of DNA damage by free diffusion

The repair of DNA damage requires the ordered recruitment of many different proteins that are responsible for signaling and subsequent repair. A powerful tool for studying the orchestrated accumulation of these proteins at damage sites is laser microirradiation in live cells, followed by monitoring of the accumulation of the fluorescently labeled protein in question. Despite the widespread use of this approach, there exists no rigorous method for characterizing this process quantitatively. Here we introduce a free diffusion model that explicitly accounts for the unique topology of individual nuclei and quantitatively describes the accumulation of two test proteins, poly-ADP-ribose polymerases 1 and 2. Application of our model to other proteins will yield novel insights into the timing and mechanism of DNA repair.

biochemistry↗

bioRxiv2017

Integrative Deep Models for Alternative Splicing

Advancements in sequencing technologies have highlighted the role of alternative splicing (AS) in increasing transcriptome complexity. This role of AS, combined with the relation of aberrant splicing to malignant states, motivated two streams of research, experimental and computational. The First involves a myriad of techniques such as RNA-Seq and CLIP-Seq to identify splicing regulators and their putative targets. The second involves probabilistic models, also known as splicing codes, which infer regulatory mechanisms and predict splicing outcome directly from genomic sequence. To date, these models have utilized only expression data. In this work we address two related challenges: Can we improve on previous models for AS outcome prediction and can we integrate additional sources of data to improve predictions for AS regulatory factors. We perform a detailed comparison of two previous modeling approaches, Bayesian and Deep Neural networks, dissecting the confounding effects of datasets and target functions. We then develop a new target function for AS prediction and show that it significantly improves model accuracy. Next, we develop a modeling framework to incorporate CLIP-Seq, knockdown and over-expression experiments, which are inherently noisy and suffer from missing values. Using several datasets involving key splice factors in mouse brain, muscle and heart we demonstrate both the prediction improvements and biological insights offered by our new models. Overall, the framework we propose offers a scalable integrative solution to improve splicing code modeling as vast amounts of relevant genomic data become available.\n\nAvailability: code and data will be available on Github following publication.

bioinformatics↗

bioRxiv2017

Ancient antagonism between CELF and RBFOX families tunes mRNA splicing outcomes

Over 95% of human multi-exon genes undergo alternative splicing, a process important in normal development and often dysregulated in disease. We sought to analyze the global splicing regulatory network of CELF2 in human T cells, a well-studied splicing regulator critical to T cell development and function. By integrating high-throughput sequencing data for binding and splicing quantification with sequence features and probabilistic splicing code models, we find evidence of splicing antagonism between CELF2 and the RBFOX family of splicing factors. We validate this functional antagonism through knockdown and overexpression experiments in human cells and find CELF2 represses RBFOX2 mRNA and protein levels. Because both families of proteins have been implicated in the development and maintenance of neuronal, muscle, and heart tissues, we analyzed publicly available data in these systems. Our analysis suggests global, antagonistic co-regulation of splicing by the CELF and RBFOX proteins in mouse muscle and heart in several physiologically relevant targets including proteins involved in calcium signaling and members of the MEF2 family of transcription factors. Importantly, a number of these co-regulated events are aberrantly spliced in mouse models and human patients with diseases that affect these tissues including heart failure, diabetes, or myotonic dystrophy. Finally, analysis of exons regulated by ancient CELF family homologs in chicken, and Drosophila suggests this antagonism is conserved through evolution.

genomics↗