Shiv Nadar University · PhD in Computer Science

Harshit Singh

AI for Drug Discovery · Protein Language Models · Explainable ML

I design sequence-first machine learning systems for protein and antibody binding affinity prediction, with a focus on models that are accurate, efficient to train, and interpretable enough to support scientific decisions. My recent work adapts ESM-2 with parameter-efficient fine-tuning, then uses attribution methods to test whether the model is learning biophysically meaningful interaction signals.

I am a Research Scholar at Shiv Nadar University, Delhi-NCR, and a Research Ambassador for Springer Nature India. I work across protein language models, antibody-antigen modeling, active learning, and explainable AI for drug discovery.

Research profile
Current focus
Explainable affinity prediction for protein and antibody interactions
Core methods
ESM-2, LoRA, metric learning, gated attention, Integrated Gradients
Applications
Antibody triage, protein-protein interaction modeling, ADMET prediction, molecular active learning
Portrait of Harshit Singh
Harshit Singh · PhD Researcher in AI for Drug Discovery

Research Snapshot

0.89 Pearson r on PPB-Affinity with BALM-PPI
<1% trainable parameters using LoRA-based fine-tuning
67.3% mean Recall@2% for GP active learning vs. 8.9% random (ACML 2026)
3-stream chain-aware antibody architecture for heavy, light, and antigen sequences

My active-learning benchmark for antibody-antigen binding affinity has been accepted at ACML 2026 (Main Track): a Tanimoto-kernel Gaussian Process over ESM-2 embeddings reaches 67.3% mean Recall@2% versus 8.9% for random sampling under a fixed 10% budget. Recent work also includes NeurIPS 2026 workshop papers on LLM proposers for molecular search (LEAF) and representation effects in molecular active learning (TabAL), plus GRASP, a graph Transformer for molecular property prediction.


Flagship Projects

Protein affinity

BALM-PPI

Fine-tuned protein language models for binding affinity prediction through metric learning, with residue-level Integrated Gradients explanations.

Antibody modeling

AbAffinity

A sequence-only, chain-aware architecture that models heavy chain, light chain, and antigen interactions without requiring solved structures.

Active learning

Antibody Active Learning (ACML 2026)

Gaussian-process active learning across eight AbBiBench assays, with attributions validated against crystal structures.

Molecular property prediction

GRASP

A 93.5M-parameter graph Transformer pretrained on ZINC20 and adapted on 642 ChEMBL assays, reaching mean MAE 0.374 on 23 OpenADMET endpoints.

Protein affinity

SIMBA

Structure-free prediction of absolute ΔG and mutation effects from frozen ESM-2 with an interface memory and Pairformer-style refinement.

See the full publication record and project details on the research page.

I also keep a short lab notebook for research notes, implementation decisions, and work-in-progress ideas.


Experience

Research Scholar — AI in Drug Discovery
Shiv Nadar University, Delhi-NCR
Research Ambassador
Springer Nature India
Assistant Professor
Noida International University, Noida (UP)
Assistant Professor
Institute of Engineering & Technology, Dr. R.M.L. Avadh University, Ayodhya
P.G.T., Computer Science
Sunbeam School, Ayodhya
Subject Matter Expert, Computer Science
Innovalance Learning System (Remote)

Education

PhD in Computer Science (Pursuing)
Shiv Nadar University, Delhi-NCR
Master of Computer Applications — Gold Medal
IET, Dr. R.M.L. Avadh University, Ayodhya
Bachelor of Computer Applications — Rank 1
IET, Dr. R.M.L. Avadh University, Ayodhya

Technical toolkit

Languages

Python (expert), SQL, C/C++, Java, JavaScript, HTML/CSS, LaTeX

ML & deep learning

PyTorch, Hugging Face (Transformers / PEFT), Scikit-learn, Keras, CatBoost

Bio & cheminformatics

ESM-2, RDKit, OpenBabel, ANARCI (antibody numbering), NGL Viewer

Advanced AI techniques

LoRA, metric learning, gated cross-attention, Integrated Gradients (XAI), self-supervised learning, ensemble learning, active learning

Developer tools

Git, Docker, GCP, Linux/Bash, VS Code, Weights & Biases

Data science

Pandas, NumPy, Matplotlib, SciPy, Captum


Awards

Harshit Singh receiving the MCA Gold Medal
MCA Gold Medal, Dr. R.M.L. Avadh University.
Harshit Singh at Springer Nature India Research Conclave 2025
Springer Nature India Research Conclave 2025.
Open CICAG Distillate Winter 2025-26 PDF
RSC-CICAG / BMCS bursary award listing.