Owing to our team's diverse skill set, we are actively involved in several projects associated to numerous aspects of human health.
Our current projects assimilate a diversity of human health concerns, such as challenges related to reproductive health, including yet not limited to prediction of IVF outcomes, determinants of IVF success, implications of vaginal microbiome and endometrium on host's reproductive health, preimplantation embryonic development, and medical diagnosis from clinical data. Significant contributions have also been made towards Vitiligo research by incorporating a co-expression network-based approach in our research. With a data-driven modus operandi, we have formed close collaborations with hospitals and research facilities to provide hybrid intelligence using our academic understanding and computational skills. We believe in creativity, innovation and brilliance and we abide by academic zeal. For suggestions and comments, the group leader can be contacted directly.
Bioinformatics in Reproductive Health and Medicine
Artificial Intelligence to improve current In-vitro Fertilization methodologies (Funded by DBT-2020)
Even though recent technological evolution in assisted reproduction has resulted in several new advances in In Vitro Fertilization procedure, the live birth rate remains comparatively low. With the aim of reducing socio-economic burden from the infertile couple, we intend to use artificially intelligence to provide more explainable and reliable solutions in the field of reproductive health and will assist the clinicians in decision making. This research study might be the very beginning of personalized prediction system that is specifically designed for Indian population.
Machine Learning/Artificial Intelligence to study infertility
A recent report suggests that approximately 48 million couples worldwide and 1 in 4 couples in developing countries experience infertility due to many distinct medical problems and lifestyle behaviors, as well as an unknown combination of them. We aim to predict and perform the accurate identification of causes of infertility and the underling networks of molecular entities using complex network construction, machine learning techniques and artificial intelligence. This study can help unraveling unknown causes of infertility, provide therapeutic targets for restoring fertility and contribute towards current assisted reproductive techniques.
Machine Learning/Artificial Intelligence to study human preimplantation embryo development
Preimplantation embryonic development (PED) of mammals covers the period from fertilization to implantation in the endometrium. However, species-specific differences like genome activation, the gene expression patterns, chromosome frequency poor segregation and patterns of epigenetic modifications can limit the extrapolation of some discoveries to the human preimplantation embryo development. Our goal is to shed more light on the complex effects about the functional pathways and underlying mechanisms of these molecular events and to provide a clearer understanding of human pre-implantation development at cellular and molecular resolution for prediction, using ML/AI approaches.
Transcriptomics to diagnose and better understand infertility caused due to unexplored and unidentified features of the Endometrium.
Embryo implantation failure is one of the major causes of infertility which is affected by both the quality of the embryo, and characteristics of the endometrium. Using Transcriptomics, we intend to make significant contributions towards diagnosing the underlying reasons behind infertility caused due to endometrium characteristics accurately. This study could help the couples enduring unexplained infertility to undergo successful embryo implantation and conceive a child with or without IVF, especially in developing country like ours.
Metagenomics to understand the contribution of vaginal microbiome to host’s reproductive health
Since the vaginal microbiome functions as preliminary defence against infections and dangerous pathogens that may invade the female reproductive tract, any fluctuation or vaginal microbial dysbiosis can lead to severe infections, disorders, and pregnancy related complications such as Pre-Term Births, Preeclampsia, etc. Using metagenomics methods and methodologies, our goal is to acquire a better understanding of the association and correlation between the vaginal microbiome and the reproductive health of the host, and the underlying processes that lead to microbial dysbiosis and related disorders. This study can help clinicians make predictions about infections or pregnancy related complications that a woman might go through based on the composition of her vaginal microbiome.
Genome Scale Metabolic Network Reconstruction
The reconstruction of a genome-scale metabolic network allows researchers to have a deeper insight at an organism's molecular processes. In these models, both genome and molecular physiology are interconnected. We perform network reconstructions to examine metabolic pathways in the context of the entire network by dissecting them into their fundamental processes, enzymes, and reactions.
GEMS for Generally Recognized as Safe (GRAS) organisms for optimization of metabolite productions used in industry
We can discover critical knowledge about the genetic makeup of various metabolic characteristics and the overall metabolic biochemical processes in a GRAS-microorganism through the evaluation and validation of Genome scale metabolic network and their reconstructions. Our aim would be to produce high-value deliverables, whether those deliverables are medically relevant, such as pharmaceuticals, high-value chemical intermediate products, or biotechnological products, such as biofuels.
GEMS for pathogenic microbes to identify drug targets and study gene essentiality
As the first step towards deciphering the disease’s underlying complex processes, we can use genome scale metabolic network reconstructions to investigate how a parasite operates in a host cell, the metabolites required for the organism’s survival and proliferation, and the essential genes that are needed for the cell to remain pathogenic and virulent. Our next step would be to utilize the reconstructed model’s predictions to identify potential biomarkers or therapeutic drug targets to improve on existing drug delivery systems.
Integration of transcriptomic data with Genome-scale Metabolic Models for cancer
Transcriptomic data can be integrated into the network topology using genome-scale metabolic models, allowing for a better physiological and molecular understanding of cancer. A variety of methods can be used to predict how different transcriptional modifications are translated into alterations in protein-encoding genes, their functions and how these altered genes/proteins interact with other biologically active compounds, leading to the formation and progression of malignant or cancerous tumours.
Vitiligo Pathogenesis
Vitiligo is a chronic, asymptomatic disorder that causes a loss of patches of skin pigment and colour.
Our Vitiligo Information Resource 1.0 (VIRdb 1.0) provides a complete platform dedicated to vitiligo, integrating both the drug-target and systems approach, along with consolidated protein and gene-level information, and potential therapeutic leads. VIRdb 1.0 also showcases a genetic interaction network of the many expressed genes in Lesional, Peri-Lesional and Non-Lesional Vitiligo compared to healthy controls, in order to completely capture the genetic complexity of Vitiligo.
VIRdb 2.0, a comprehensive set of differentially co-expressed genes involved in crosstalk events between Vitiligo and associated autoimmune disorders (Multiple Sclerosis, Psoriasis and Rheumatoid Arthritis) is presented and incorporated with Vitiligo-related multi-omics datasets (connected to SwissProt, KEGG, GeneCards, NPASS and STRING). We anticipate that researchers and clinicians working on vitiligo drug development will benefit from VIRdb 2.0.