The availability of huge omics data from genome projects and high-throughput technology (next-generation sequencing and microarray) has brought a great challenge to understand the complexity of biological processes and disease mechanisms in eye research. We seek an agile and predictive understanding of how genetic variants that results in eye diseases, including ocular cancer. We write algorithms and pipelines to get critical answers faster from NGS data. We also focus on non-coding RNA expression and its regulatory role in eye diseases by integrating data from NGS and the public. We have a reliable infrastructure and framework comprised of LINUX and Windows-based servers and desktop workstations, which allow us to integrate high throughput data and study them at the systems level. Bioinformatics centre is highly interdisciplinary, at the interface of Biology, computational biology and Informatics.
We primarily focus on developing bioinformatics methods of NGSomics data analysis and the role of small noncoding RNAs as biomarkers for vision biology and eye diseases. This includes whole genome, exome, transcriptome next-generation sequencing data analysis in identifying molecular targets for ocular cancer and diseases, the small-RNA seq analysis of human with focus on profiling micro-RNAs and identifying their regulatory role in eye diseases, comparative genome analysis to identify virulence and drug resistance mechanisms in ocular pathogens and structure-based approaches of non-synonymous variants (nsSNVs) to understand the molecular mechanisms for pathogenicity. As part of the multidisciplinary nature of our field, we work on these projects in close collaboration with wet-lab scientists. The following gives a short summary of a few selected research projects.
Develop automated pipelines and infrastructure to detect pathogenic variants from exome data of eye disease patients. The exome sequencing studies primarily aim at the discovery of single nucleotide variants (SNVs), and small insertions and small deletions (INDELs) of coding region that is about 85 % of mutations among all the genetic variations. Exome sequencing method has been widely used to elucidate the genetic causes of many eye diseases, starting from single gene disorders and moving on to more complex genetic eye disorders, including complex traits and cancer. Although the exome sequencing has demonstrated identifying clinical variants, bioinformatics challenges are being faced as the current bottleneck in exome/genome methods shifted from sequence generation to data management and analysis. Using our in-house automated pipeline, we can find almost all variants within the targeted region of the genomes. We apply a series of filters to identify the potential disease-causing variants. We are developing a statistical model to specifically filter eye disease causing variants.
Recent advances in next generation sequencing (NGS) methods have brought a paradigm shift in discovering eye disease-associated genetic variants from linkage and genome-wide association studies to NGS-based genome/exome studies. Mainly, whole-exome sequencing (WES) is now made as a viable approach to uncover the pathogenic variants for both Mendelian and complex eye diseases with a limited number of probands. WES, focuses on only the protein-coding sequence of the human genome, is become a powerful tool with many advantages in the research setting, and moreover, is now being implemented into the clinical diagnostic arena. Nevertheless, the identification of pathogenic variants remains a great challenge. Pathogenic variant prioritization using simple heuristic filtering approaches and functional implications of variants misses the true positives. Using our in-house automated pipeline, we can find almost all variants within the targeted region of the genomes. We use stringent filtering methods and machine learning methods to prioritize pathogenic variants for eye diseases. We further aim to achieve a variant prioritized model to correctly filter eye disease-causing variants for Mendelian and complex eye diseases. In tandem, we work on translational genomics of ocular cancer including Retinoblastoma and ocular lymphoma.
Comparative genomics approach of ocular isolates from keratitis patients with different clinical outcomes used to better understand the infection, genome-wide identification of genetic features responsible for multiple virulence and multidrug-resistant (MDR) mechanisms. Microbial keratitis due to either fungus or bacteria is a major cause of blindness in India. The bacterial keratitis, often caused by Pseudomonas aeruginosa and methicillin-resistant staphylococcus aureus (MRSA) at Aravind Eye hospital, Madurai, majority of times in spite of adequate medical management the ulcer does not heal and may require a corneal transplant. Pseudomonas aeruginosa can cause a most severe keratitis, carrying a wide array of virulence factors that contribute pathogenesis. Keratitis pathogenesis is a complex process, where in several virulence factors has been implicated including Cell-associated structures such as type IV pili and flagella, slime polysaccharide, proteases such as elastase B (LasB), alkaline protease, protease IV and Pasp and exotoxins. Also, clinical isolates of Pseudomonas often exhibit multiple resistances to antibiotics. On the other hand, studies have shown an increase in the pervasiveness of ocular MRSA infections. Here, comparative genomics through de novo assembly of NGS data of ocular isolates of P. aeruginosa carrying specific features compare to other strains suggesting that they may be adapted with these features to cause eye infections. We are using same approach to study MRSA ocular pathogens and further studying the link between MDR genotypes and clinical outcome or virulence factors.
Computational strategies to analyze intrinsic relationships among dysregulated miRNA and their target interactions in infection and explore their regulatory role and as potential biomarkers. MicroRNAs are a novel group of non-coding small RNAs that post-transcriptionally control gene expression by promoting either degradation or translational repression of target messenger RNA. They are implicated in a large variety of physiological and pathophysiological processes. Levels of miRNAs in the serum of humans have been shown to be stable, reproducible, consistent amongst healthy individuals and change during pathophysiology, and their presence in ocular fluids allowing them to be of potential value as clinical biomarkers of eye disease. We focus on miRNAs regulatory role from dysregulated human miRNAs identified through small-RNA sequencing in glaucoma, diabetic retinopathy and fungal keratitis (IOVS, 2015). In addition, we are looking at their regulatory role in maintenance of limbalstemness with wet lab scientist.
Use structural bioinformatics to investigate functional impact of non-synonymous single nucleotide variants and its association with genetic eye diseases. The advent of NGS identified several new candidate genes for eye diseases. Variations in the same gene can cause very different eye diseases (pleiotropy), on the other hand, single disease can be genetically heterogeneous. Thus, a detailed underlying molecular mechanism is needed to understand this complexity. Non-synonymous single nucleotide variants (nsSNVs) in the coding region of the protein is of critical importance to understand the molecular mechanisms of the disease and to clarify the association between patient-specific variants and disease phenotype. We use simple analytical strategy using protein structure to predict the pathogenicity of nsSNVs and investigate their functional impact that may leads to eye disease phenotypes.
Develop targeted analysis pipeline for RB1 gene and other cancer associated genes to identify pathogenic variants for the molecular diagnosis of Retinoblastoma. Accurate identification of pathogenic variants in a reduced time is very important for diagnosis, confirmation, genetic counselling, risk assessment, and carrier screening of Retinoblastoma (RB) patients and their family members. However, genetic analysis of heterogeneous spectrum of variants in RB1 gene and other cancer associated genes is no trivial task and essentially requires comprehensive approach. Target enrichment followed by next-generation sequencing offer a time-efficient and accurate approach for the molecular diagnosis of many eye diseases. However, identifying pathogenic variants, moreover, spectrum variations, is challenging as data analysis requires several bioinformatics tools, data management, pathogenic variant filtering and reporting. Our in-house bioinformatics pipeline can identify heterogeneous spectrum of RB1 gene variants including SNVs, InDels and CNVs for the molecular diagnosis of RB. We develop the automated targeted analysis pipeline to both germline and somatic variants for ocular cancer gene panel.
The AMRF Biocomputing Center (ABC) provides a core computational facility with a reliable infrastructure and framework equipped with Dell T630 Server (With Ubuntu 14.04) and HP DL580R07 (E7) CTO Server, three Dell workstations and five Intel i7-3370 3.5GHz workstations. It is a multidisciplinary research environment that provides to customize data analysis tailored to the needs of individual research projects across all the research groups and extend this service to others on mutually acceptable terms. In addition, it helps to train manpower by way of workshops and short training courses.
Aravind Medical Research Foundation (AMRF), Madurai
Science and Engineering Research Board (SERB), Govt. of India.
Department of Biotechnology (DBT), Govt. of India
Next-generation sequence data processing and analysis: The resource has developed processing and analysis pipelines as below for illumina and ion data. The input to the NGS pipeline is either raw reads from the sequencing machines or mapped reads from alignment software.
Microarray data processing and analysis: This includes background correction, normalization, summarization, quality control, detecting differentially expressed genes, and correlation of gene expression with phenotypes or clinical variables.
Bioinformatics: The resource is available to help investigators with bioinformatics analysis such as pathway and gene function enrichment analysis and gene network analysis.
Database: Services include design and implementation of interactive web applications as well as the underlying database back-ends. Support is available for investigators with problems concerning data acquisition, management, and analysis.