Biology subjects

Subramanian, U.

Publications and source records attributed to Subramanian, U..

2 recordsLinked to original sources

bioRxiv2020

Identification of Dysregulated Pathways and key genes in Human Retinal Angiogenesis using Microarray Metadata

Retinal angiogenesis is a common neovascularization mechanism that causes severe irreversible vision loss in the number of retinal diseases worldwide. Patients often do not respond to the current anti-angiogenic therapies and have a vision loss. Understanding the various angiogenic pathways and factors involved in the pathogenic mechanism is vital for disease management. In this study, to identify dysregulated angiogenic pathways and specific angiogenic factors involved in vision-threatening diseases namely proliferative diabetic retinopathy (PDR), retinopathy of prematurity (ROP) and neovascular age-related macular degeneration (nAMD), we downloaded microarray metadata of samples and obtained the differentially expressed genes (DEGs) in all the disease and each disease samples compared to controls. Subsequently, we performed Gene Set Enrichment (GESA) analysis for pathways, a protein-protein interaction (PPI), and angiome network analysis using R and Cytoscape software. We identified highly enriched dysregulated pathways that were neuroactive ligand receptor interaction and cytokine-cytokine receptor interaction. The angiogenic-associated DEGs were predominately related to the cytokine-cytokine receptor interaction pathway, which we further confirmed with RNA-seq data of PDR samples. Together, our analysis of these data elucidated the molecular mechanisms of retinal angiogenesis and provided potential angiogenic targets for therapeutics.

bioinformatics↗

bioRxiv2018

Performance Assessment of Variant Calling Pipelines using Human Whole Exome Sequencing and Simulated data

The whole exome sequencing (WES) is a time-consuming technology in the identification of clinical variants and it demands the accurate variant caller tools. The currently available tools compromise accuracy in predicting the specific types of variants. Thus, it is important to find out the possible combination of best aligner-variant caller tools for detecting SNVs and InDels separately. Moreover, many important aspects of InDel detection are not overlooked while comparing the performance of tools. One such aspect is the detection of InDels with respect to base pair length. To assess the performance of variant (especially InDels) caller in combination with different aligners, 20 automated pipelines were developed and evaluated using gold reference variant dataset (NA12878) from Genome in a Bottle (GiaB) consortium of human whole exome sequencing. Additionally, the simulated exome data from two human reference genome sequences (GRCh37 and GRCh38) were used to compare the performance of the pipelines. By analyzing various performance metrices, we observed that BWA and Novoalign aligners performed better with DeepVariant and SAMtools callers for detecting SNVs, and with DeepVariant and GATK for Indels. Altogether, DeepVariant with BWA and Novoalign performed best. Further, we showed that merging the top performing pipelines improved the accurate variant call set. Collectively, this study would help the investigators to effectively improve the sensitivity and accuracy in detecting specific variants.

bioinformatics↗