Biology subjects

SINGH, R. K.

Publications and source records attributed to SINGH, R. K..

8 recordsLinked to original sources

bioRxiv2026

Active Learning framework for Antibody Binding Affinity Prediction

Antibody engineering campaigns must prioritise variants from large combinatorial libraries under limited assay budgets. We present a Gaussian process (GP) active-learning benchmark for antibody-antigen binding affinity prediction across eight public AbBiBench assays spanning influenza hemagglutinin antibodies CR6261 and CR9114 and AAYL-series scFv libraries against the SARS-CoV-2 Spike HR2 peptide. The framework embeds variable heavy and light (VH/VL) sequences with ESM-2 650M, trains GP surrogates over compressed protein-language-model representations, and evaluates kernel, acquisition, representation, classical featurisation, and labelling-budget choices under a unified budget of 10\% of each full pool with a label-blind seed and a fixed held-out evaluation split. Across the benchmark, Tanimoto+UCB has the highest mean Recall@2\% point estimate among tested kernel-acquisition configurations, achieving \textbf{57.8\%} mean Recall@2\% across five random seeds (68.3\% when scored against the acquisition-eligible pool), compared with 9.6\% for random selection -- a gap that is significant by a paired Wilcoxon test across datasets ($p=0.0078$); its point estimate is not significantly different from RQ+UCB or Tanimoto+EI at five seeds. We adopt the continuous Tanimoto kernel of \citet{gessner2024}. The same configuration recovers the largest fractions of top binders in the influenza libraries (74.2\% on \texttt{3gbn\_h1}, 82.7\% on \texttt{3gbn\_h9}, and 84.9\% on \texttt{4fqi\_h3}) while exposing harder AAYL combinatorial landscapes where recall remains lower and more variable. An additional three-seed representation ablation, run under the original protocol and not repeated after the evaluation corrections, ranked ESM-2 650M highest (mean Recall@2\% 0.673 $\pm$ 0.28) among ProtBert, AntiBERTy, AbLang2, ProGen2, and classical encodings. On BindingGYM, the same configuration matches or exceeds ALLM-Ab on two of three datasets using the 100-sample held-out test metric. Structure-mapped Integrated Gradients and KernelSHAP analyses recover CDR-region attribution patterns consistent with contact-derived paratope geometry, though threshold-free (AUROC) analysis shows this qualitative agreement is not matched by strong residue-level discrimination. These results support Tanimoto-kernel GP active learning as a simple, reproducible strategy for label-efficient antibody affinity optimisation with qualitative structural validation.

bioinformatics↗

bioRxiv2026

Do LLM Proposers Improve Budgeted Search for Ionizable Lipids?

Efficient lipid design under limited experimental budgets is a central challenge in computational chemistry and drug discovery. We introduce LEAF (Lipid Evolutionary Adaptive Framework), a matched-budget protocol for testing whether language-model proposers improve combinatorial molecular search relative to classical baselines. The study uses a 30,044-candidate ionizable-lipid space, a frozen Gaussian-process oracle fitted to experimental potency labels, and a separate active-learning surrogate. Four language models evaluated over eight seeds and three splits show no significant improvement over genetic-algorithm or random-search baselines. For Llama 8B, 79.1% of acquisition-selected queries repeat previously evaluated archive members. Removing duplicates reduces the performance gap on some splits but increases proposal cost and can prevent budget exhaustion. Larger models reduce duplication without establishing superiority or equivalence. Exploratory scored prompts and stricter novelty controls yield positive median differences against the genetic algorithm, yet wide intervals and shared-seed dependence preclude stronger conclusions. A validated single-assay reanalysis preserves the chemotype-cold deficit at a smaller numerical scale, and exploratory expected hypervolume improvement does not reverse the result. Follow-up checks address whether proposers genuinely use injected evidence. These checks include a zero-shot leakage probe, rank-only and chain-of-thought-rationale prompts, a score-shuffle perturbation, an evidence-conflict probe, and a genuine reasoning-mode proposer. They show measurable engagement with supplied information without a corresponding performance gain. LEAF isolates proposal quality as the experimental variable and documents a concrete failure mode of zero-shot LLM proposers under controlled budgets.

bioinformatics↗

bioRxiv2026

Probing isn't Causality in Molecular Models

Linear probes establish whether a chemical variable can be recovered from a molecular representation; they do not establish how that variable contributes to subsequent predictions. We replace a concept-associated scalar component at an intermediate block and measure the change through a separately fitted downstream outcome readout. Comparisons with random directions and architecture-matched random encoders distinguish recoverability, intervention selectivity, and a pretraining-associated selectivity advantage. In Uni-Mol, block-12 total-distortion edits predict the observed energy-change sign for 97.6\% of held-out pairs, with Spearman $\rho=0.787$. Structure-clustered intervals for the primary direction and encoder contrasts exclude zero. A stricter 20-direction sensitivity analysis does not survive correction across six cells. A post-review analysis finds substantial concept--outcome alignment; intervention effects persist after removal of the same-block outcome direction, but this does not establish that subsequent encoder computation is necessary. In MolCLR, pretraining improves dipole recovery by $0.080\,R^2$, whereas concept-minus-random sign-accuracy contrasts are $0.029$ and $-0.015$ at the two tested blocks. Their paired uncertainty is unresolved, so they do not establish an absence of selective effects. We propose an operational evidence ladder whose conclusions are conditional on the concept, intervention, control family, and fitted readout.

bioinformatics↗

bioRxiv2026

Antibody-Antigen Affinity Prediction with Chain-Aware Protein Language Modeling

MotivationAntibody-antigen affinity determines which antibodies advance in therapeutic discovery, repertoire analysis and affinity maturation, but experimental measurements are sparse relative to the scale of sequence libraries. Structure-based predictors can exploit interface geometry when reliable complexes are available, yet early discovery often requires ranking many heavy-light chain pairs against antigens for which no complex structure exists. Existing sequence-based models are scalable, but frequently compress heavy and light chains into a single antibody representation or concatenate antibody and antigen features obscuring the chain-specific and epitope-specific signals that drive binding. ResultsWe present AbAffinity, a sequence-only chain-aware three-stream architecture that maintains heavy chain, light chain and antigen as distinct streams. It integrates frozen ESM-2 embeddings with heavy-chain CDR-focused pooling, heavy-light self-attention, adaptive fusion gating and gated cross-attention, training only a compact interaction module. On the SAAINT-DB benchmark, AbAffinity achieves strong predictive performance under ten-fold cross-validation and maintains robust accuracy on novel antigens. It consistently outperforms recent sequence-based models across external benchmarks including SAbDab, AB-Bind and SKEMPI 2.0. Ablation studies highlight the contributions of chain-specific representations, CDR-focused pooling and the gated interaction pathway. Integrated Gradients attributions recover known paratope and epitope residues at structurally validated interfaces. AbAffinity provides a lightweight, explainable sequence-first framework for antibody triage and prioritisation when structural information is limited or unavailable.

bioinformatics↗

bioRxiv2026

Explainable protein-protein binding affinity prediction via fine-tuning protein language models

Protein-protein interactions underpin virtually every aspect of cellular life, and the precise quantification of their binding affinity is fundamental to understanding immune recognition, disease mechanisms, and the rational design of therapeutic antibodies. Yet predicting binding affinity at scale remains an unsolved challenge: reliable experimental assays are low-throughput and expensive, while computational methods that depend on three-dimensional complex struc-tures cannot be applied to the vast majority of clinically relevant targets where structural data are absent. Here we present BALM-PPI, a framework that predicts protein-protein binding affinity from amino acid sequence alone. Both proteins are encoded by a protein language model trained on evolutionary sequence data and projected into a shared representational space, where their distance directly reflects binding strength. Fine-tuning this protein language model requires updating fewer than 1% of its parameters, and we show that this targeted adaptation steers the model toward interface-relevant sequence signals rather than spurious background correlations. On a curated benchmark of over 12,000 protein complexes, BALM-PPI matches or exceeds the accuracy of structure-based methods and retains predictive power for proteins with less than 30% sequence identity to the training set. Using only a subset of project-specific assay data, BALM-PPI outperforms a recent method trained on three times the data, suggesting that the model has already encoded the underlying interaction signals and requires only minimal supervision to specialise to a new target. BALM-PPI further provides residue-level attribution maps that pinpoint the amino acid positions driving each affinity prediction, consistently re-covering experimentally validated interaction hotspots across enzyme-inhibitor, signalling, and antibody-antigen systems without any structural input during training. This allows predictions to be cross-validated against structural and mutagenesis evidence, providing a mechanistic basis for candidate shortlisting ahead of experimental follow-up. BALM-PPI is freely accessible via an interactive web server.

bioinformatics↗

bioRxiv2025

Explainable Active Learning Framework for Ligand BindingAffinity Prediction

Active learning (AL) prioritises which compounds to measure next for protein-ligand affinity when assay or simulation budgets are limited. We present an explainable AL framework built on Gaussian process regression and assess how molecular representations, covariance kernels, and acquisition policies affect enrichment across four drug-relevant targets. Using recall of top active compounds, we find that dataset identity--the targets chemical landscape--sets the performance ceiling, while method choices modulate outcomes rather than overturn them. Fingerprints with simple Gaussian process kernels provide robust, low-variance enrichment, whereas learned embeddings with non-linear kernels can reach higher peaks but with greater variability. Uncertainty-guided acquisition consistently outperforms random selection, yet no single policy is universally optimal; the best choice follows structure-activity relationship (SAR) complexity. To enhance interpretability beyond black-box selection, we integrate SHapley Additive exPlanations (SHAP) to link high-impact fingerprint bits to chemically meaningful fragments across AL cycles, illustrating how the models attention progressively concentrates on SAR-relevant motifs.

biophysics↗

bioRxiv2020

A real-time, transient kinetic study of Drosophila melanogaster Dicer-2 elucidates mechanism of termini-dependent cleavage of dsRNA.

Drosophila melanogaster Dicer-2 (dmDcr-2) differentially processes dsRNA with blunt or 2 nucleotide 3-overhanging termini. We investigated the transient kinetic mechanism of these reactions using a rapid reaction stopped-flow technique and time-resolved fluorescence spectroscopy. We found that ATP binding to dmDcr-2s helicase domain impacts the kinetics of dsRNA binding and dissociation in a termini-dependent manner, emphasizing the termini-dependent discrimination of dsRNA on a biologically-relevant time-scale. ATP-hydrolysis mediates local unwinding of dsRNA, and directional translocation on unwound single-stranded RNA, which is concurrent with a slow rewinding prior to dsRNA cleavage. Time-resolved fluorescence anisotropy reveals a nucleotide-dependent change in conformational dynamics of the helicase and Platform*PAZ domains in the nanosecond timescale that is correlated with termini-dependent dsRNA cleavage. Our study delineates kinetic events and transient intermediates for a Dicer-catalyzed reaction, thus establishing a framework for understanding other Dicers and how accessory factors modulate the reaction.

biophysics↗

bioRxiv2019

DMENet: Diabetic Macular Edema Diagnosis using Hierarchical Ensemble of CNNs

Diabetic Macular Edema (DME) is an advanced stage of Diabetic Retinopathy (DR) and can lead to permanent vision loss. Currently, it affects 26.7 million people globally and on account of such huge number of DME cases and limited number of ophthalmologists, it is desirable to automate the diagnosis process. Computer-assisted, deep learning based diagnosis will help in early detection, following which appropriate medication can help to mitigate the vision loss. Method: In order to automate the screening of DME, we propose a novel DMENet Algorithm which is built on the pillars of Convolutional Neural Networks (CNNs). DMENet analyses the preprocessed color fundus images and passes it through a two-stage pipeline. The first stage detects the presence or absence of DME whereas the second stage takes the positive cases and grades the images based on severity. In both the stages, we use a novel Hierarchical Ensemble of CNNs (HE-CNN). This paper uses two of the popular publicly available datasets IDRiD and MESSIDOR for classification. Preprocessing on the images is performed using morphological opening, Gaussian kernel and the dataset is augmented to solve the class imbalance problem. Results: The proposed methodology achieved an Accuracy of 96.12%, Sensitivity of 96.32%, Specificity of 95.84%, and F-1 score of 0.9609. Conclusion: These excellent results establishes the validity of the proposed methodology for use in DME screening and solidifies the applicability of HE-CNN classification technique in the domain of bio-medical imaging.

bioengineering↗