Research

Virtual screening (VS) aids in prioritizing unknown bio-interactions between compounds and protein targets for empirical drug discovery. In standard VS exercise, roughly 10% of top-ranked molecules exhibit activity when examined in biochemical assays, which accounts for many false positive hits, making it an arduous task. Attempts for conquering false-hit rates were developed through either ligand-based or structure-based VS separately; however, nonetheless performed remarkably well. Here, we developed an advanced VS framework —automated hit identification and optimization tool (A-HIOT)— comprises chemical space-driven stacked ensemble for identification and protein space-driven deep learning architectures for optimization of an array of specific hits for fixed protein receptors. In conclusion, advantageous features impeded in A-HIOT is making a reliable approach for bridging the long-standing gap between ligand-based and structure-based VS in finding the optimized hits for the desired receptor.

Code available for Automated Hit identification and Optimization Tool (A-HIOT) Gitlab

Neeraj Kumar, Vishal Acharya*, Machine intelligence-driven framework for optimized hit selection in virtual screening Journal of Cheminformatics 14, 48 (2022). https://doi.org/10.1186 Article (IF=8.48)

Host plants are continually competing with pathogens and vice-versa for survival and development. Such interactions in agriculturally important species can directly impact crop yield. One of the major obstacles to attaining maximum plant yield is the severity of diseases caused by multiple pathogenic organisms. To overcome this, plants display dynamic, carefully controlled transcriptome alterations in response to pathogen challenges. Both positive and negative regulators regulate plant immune signaling. It is critical to balance growth and defense responses to respond appropriately to environmental changes.Interactome mappings in diverse plant species during the past 20 years have led to the expansion of network biology premises. However, it was not enough to gain insights from a single molecular system description in plant-pathogen interactions. Therefore, to address this, we have developed such networks based on expression profile data of model plant, here, rice in response to three biotic stresses, leading to the construction of DL-based supervised Net (DLNet). This is the first ever implementation of integrated deep learning and network biology approach to understand the adaptation of plant immune genes in response to multiple pathogens using genomics data.

The code for DLNet algorithm is available GitLab

RicePathDLNet (Rice pathogen Deep Learning Network): Database for visualization software Browse

Ravi Kumar, Abhishek Khatri, Vishal Acharya*, Deep learning uncovers distinct behaviour of rice network to pathogens response. iScience, Cell Press (2022). Article (IF=6.107)

Human Cancer (Oxidative Stress)

Cancer system biology is used to elucidate mechanisms for cancer progression, but networks defining mechanisms causing resistance is less explored. Using system biology, we identified DNA damage response (DDR) hubs between G1-S and M phases that were associated with bad prognosis. The increased expression of DDR network was involved in resistance under high oxidative stress. We validated our findings by combining H2O2 induced oxidative stress and DDR inhibitors in human lung cancer cells to conclude the necessity of targeting a 'disease-causing network'. Collectively, our work provides insights toward designing strategies for network pharmacology to combat resistance in cancer research.

Meetal Sharma, Prince Anand, Yogendra Padwad*, Vivek Dogra*, Vishal Acharya*, DNA damage response proteins synergistically affect the cancer prognosis and resistance. Free radical Biology and medicine (2021) Article(IF=8.1)