Biology subjects

Srivastava, S. P.

Publications and source records attributed to Srivastava, S. P..

15 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

CLARA: Compound Library Active Learning with Reasoning Agents

Batch active learning for ligand discovery requires a recurring decision about whether to exploit high predicted affinity, reduce model uncertainty, or broaden chemical coverage. We test whether a language model can make this strategy-level decision while candidate ranking and label revelation remain controlled by a deterministic Gaussian-process environment. Across four protein targets, ten paired seeds, and a 360-compound assay budget, a GPT-5.6 Sol agent outperformed random sampling in all 40 matched campaigns. Its final top-2\% recall was 0.412 on D2R, 0.374 on Mpro, 0.448 on TYK2, and 0.683 on USP7. It was statistically tied with the strongest fixed policy on Mpro and USP7 but underperformed target-specific leaders on D2R and TYK2. The agent shifted from exploration-oriented early batches toward exploitation and produced 686 testable threshold claims with an 86.3\% atomic success rate, although 19.8\% of forecasts were not numerically scorable. In 200 prespecified information-ablation campaigns, removing uncertainty, diversity, or memory caused only small endpoint changes, while shuffling explanatory evidence reproduced the full allocation path in 39 of 40 paired campaigns. A deterministic 80\% availability constraint reduced recall on every target but preserved the target-dependent pattern of wins and losses. These results support language-model strategy control as a feasible and auditable layer.

bioinformatics↗

bioRxiv2026

GRASP: Graph Representation Learning with Assay Supervision for Molecular Properties

Molecular structures are abundant, while experimental bioactivity is sparse and distributed across assays. We introduce GRASP, a 93.5M-parameter graph Transformer that learns from these sources in sequence. GRASP first learns molecular structure through replaced-token detection on 1.54 billion ZINC20 molecule presentations, then adapts to sparse bioactivity across 642 ChEMBL assays. Across 23 cluster-held-out OpenADMET endpoints, GRASP achieves a mean MAE of 0.374, compared with 0.383 for CheMeleon, 0.408 for ECFP4-LightGBM, and 0.440 for Chemprop. ChEMBL adaptation reduces mean MAE from 0.407 to 0.374 and improves 21 of 23 endpoints, showing the value of intermediate bioactivity supervision. On 22 TDC ADMET tasks, GRASP gives a better point estimate than CheMeleon on 13 tasks. These results support staged molecular representation learning, where abundant molecular structure and sparse experimental measurements provide complementary supervision at different stages of training.

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↗

bioRxiv2025

Metabolic reprogramming by endothelial ANGPTL4 depletion protects against diabetic kidney disease

The role of cell-specific ANGPTL4 is not well known in the context of ECs, specifically in pathological angiogenesis and its relation to diabetic kidney disease. Here, we demonstrate that endothelial ANGPTL4 is required to induce a metabolic phenotype that favors mesenchymal activation in ECs and tubules in diabetic conditions. Diabetes accelerates mesenchymal activation and fibrogenesis in control mice however, the same effects were not observed in endothelial-cell specific knock out mice. This mesenchymal activation in diabetes is directly linked with pathological neovascularization, endothelial leakage, lipid and glycolysis metabolite load, de novo lipogenesis (DNL) and related mitochondrial damage, activation of the immune system, c-GAS-STING activation and transcription of pro-inflammatory cytokines. However, endothelial ANGPTL4-depleted mice had stable vessels, improved levels of lipid and glucose metabolism, suppressed levels of DNL, restored mitochondrial function, and mitigated levels of c-GAS-STING-mediated inflammation. Moreover, Inhibition of DNL, and STING via small molecule inhibitors suppressed pathological neovascularization and endothelial leakage, normalized fatty acid oxidation and reduced pathological glycolysis and de novo lipogenesis (DNL). These data demonstrate the crucial roles of endothelial ANGPTL4 in regulating pathogenic angiogenesis in the renal vasculature during diabetes.

molecular biology↗

bioRxiv2025

HUWE1 stimulates mTORC1 activity by enhancing Rheb interaction with mTORC1 and supports de novo pyrimidine synthesis

mTORC1 is a central regulator of cell growth that is directly activated by the small GTPase Rheb on the lysosomal membrane in response to growth factors. We recently reported that polyubiquitinated Rheb facilitates its interaction with mTORC1, leading to mTORC1 activation. However, the molecular mechanisms underlying the ubiquitination of Rheb and its interaction with mTORC1 for activation are not fully understood. In this study, we demonstrate that HUWE1, an E3 ubiquitin ligase, preferentially interacts with ubiquitinated Rheb and is essential for the interaction between Rheb and mTOR, as well as for mTORC1 activation. Additionally, HUWE1 is necessary for Rheb to interact with CAD, a crucial enzyme involved in de novo pyrimidine biosynthesis, and for its activation through the mTORC1-S6K1 pathway. Knocking down HUWE1 in cultured cells or in the liver tissues of mice inhibits mTORC1 activity without affecting the phosphorylation of Akt or TSC2, nor does it affect the lysosomal localization of TSC2 or mTORC1. Furthermore, HUWE1 specifically enhances the expression of CAD without influencing other enzymes involved in de novo pyrimidine synthesis and maintains UMP levels in the hepatocytes of liver tissues. These findings indicate that HUWE1 serves as a key organizer of the Rheb ubiquitin complex, amplifying mTORC1 activity and playing a vital role in stimulating de novo pyrimidine synthesis by enhancing CAD expression and activity. HighlightsO_LIHUWE1 and CAD preferentially interact with ubiquitinated Rheb. C_LIO_LIHUWE1 further ubiquitinates Rheb and supports Rheb interaction with mTORC1 and CAD. C_LIO_LIHUWE1 stimulates mTORC1 activity through its E3 ligase activity. C_LIO_LIHUWE1 also supports pyrimidine synthesis by enhancing CAD expression. C_LI

biochemistry↗

bioRxiv2025

AMPK is dispensable for physiological podocyte and glomerular functions but prevents glomerular fibrosis in experimental diabetes

AMP-activated protein kinase (AMPK) has been postulated to be crucial in regulating various renal physiology and pathophysiology processes, including energy metabolism, ion and water transport, inflammation, and hypertrophy. However, the specific roles of AMPK in the podocyte, a cell critical for maintaining glomerular filtration, have not been fully explored using genetic model animals. In this study, we generated mice lacking both AMPK 1 and 2 catalytic subunits in glomerular podocytes (pmut). Our findings revealed that, surprisingly, AMPK is dispensable for normal podocyte function. These knockout mice could live as long as their wild-type littermates without showing any pathological alterations in their glomeruli or glomerular function at two years of age. However, under type 1 diabetic conditions, the diabetic pmut mice exhibited increased lipid and collagen accumulation and an elevated expression of mesenchymal proteins in their glomeruli. They also showed more significant albuminuria compared to control diabetic mice. Under high glucose culture conditions, glomeruli isolated from pmut mice demonstrated a reduced expression of mitochondrial genes (e.g., Ndufv2) and increased leakage of mitochondrial components. Additionally, there was heightened expression of genes associated with nucleotide sensing and pro-inflammatory pathways (including mb21d2, IL-1 beta, and NF-kB). These observations suggest that while AMPK is not necessary for podocyte function in healthy kidneys, it is crucial for preventing glomerular fibrosis resulting from lipotoxicity and inflammation under diabetic conditions.

cell biology↗

bioRxiv2023

Renal Angptl4 is a key fibrogenic molecule in progressive diabetic kidney disease

Angiopoietin-like 4 (ANGPTL4) is the key protein involved in lipoprotein metabolism and has been shown to have diverse effects on tissue protection. In clinical settings, there is a reported association between higher levels of plasma Angptl4 and features of diabetic kidney disease, however, the association between kidney Angptl4 with features of diabetic kidney disease has not been well investigated. We show that both podocyte-and tubule-specific ANGPTL4 are crucial fibrogenic molecules in diabetes. Results from mRNA-array analysis in control (non-fibrotic) and diabetic (fibrotic) kidneys suggest time-dependent emergence of Angplt4 expression. Diabetes accelerates the fibrogenic phenotype in control mice but not in ANGPTL4 mutant mice. The protective effect observed in ANGPTL4 mutant mice is correlated with a reduction in the levels of pro-inflammatory cytokines, epithelial-to-mesenchymal transition, endothelial-to-mesenchymal transition and augmented fatty acid oxidation. Mechanistically, we demonstrate that podocyte-or tubule-secreted Angptl4 interacts with Integrin-{beta}1 and influences the association between dipeptidyl-4 with Integrin-{beta}1 and promotes heterodimerization of transforming growth factor-{beta} receptor 1 (TGF{beta}R1) and TGF{beta}R2 in cultured cells. This in turn results in Smad3 phosphorylation and subsequent downregulation of the expression of genes involved in fatty acid oxidation; these cumulative effects led to the activation of fibrogenic phenotypes. We demonstrate the utility of a targeted pharmacologic therapy that specifically inhibits Angptl4 gene expression in the kidneys and protects diabetic kidneys from proteinuria and fibrosis. Importantly, use of this kidney-specific targeted strategy is beneficial and does not cause any harmful effect suggesting it can be used as a novel drug molecule for treatment of diabetic kidney disease. Taken together, these data demonstrate that podocyte-and tubule-derived Angptl4 is fibrogenic in diabetic kidneys.

cell biology↗

bioRxiv2023

Loss of endothelial glucocorticoid receptor accelerates organ fibrosis in db/db mice

Endothelial cells play a key role in maintaining homeostasis and are deranged in many disease processes, including fibrotic conditions. Absence of the endothelial glucocorticoid receptor (GR) has been shown to accelerate diabetic kidney fibrosis in part through up regulation of Wnt signaling. The db/db mouse model is a model of spontaneous type 2 diabetes that has been noted to develop fibrosis in multiple organs over time, including the kidneys. This study aimed to determine the effect of loss of endothelial GR on organ fibrosis in the db/db model. Db/Db mice lacking endothelial GR showed more severe fibrosis in multiple organs compared to endothelial GR-replete db/db mice. Organ fibrosis could be substantially improved either through administration of a Wnt inhibitor or metformin. IL-6 is a key cytokine driving the fibrosis phenotype and is mechanistically linked to Wnt signaling. The db/db model is an important tool to study mechanisms of fibrosis and its phenotype in the absence of endothelial GR highlights the synergistic effects of Wnt signaling and inflammation in the pathogenesis or organ fibrosis.

physiology↗

bioRxiv2020

Endothelial SIRT3 regulates myofibroblast metabolic shifts in diabetic kidneys

Defects in endothelial cells cause deterioration in kidney function and structure. Here, we found that endothelial SIRT3 regulates metabolic reprogramming and fibrogenesis in the kidneys of diabetic mice. By analyzing, gain-of-function of the SIRT3 gene by overexpression in a fibrotic mouse strain conferred disease resistance against diabetic kidney fibrosis; while its loss-of-function in endothelial cells exacerbated the levels of diabetic kidney fibrosis. Regulation of endothelial cell SIRT3 on fibrogenic processes was due to tight control over the defective central metabolism and linked-activation of endothelial-to-mesenchymal transition (EndMT). SIRT3 deficiency in endothelial cells stimulated the TGF{beta}/Smad3-depandent mesenchymal transformations in renal tubular epithelial cells. These data demonstrate that SIRT3 regulates defective metabolism and EndMT-mediated activation of the fibrogenic pathways in the diabetic kidneys. Together, our findings show that endothelial cell SIRT3 is a fundamental regulator of defective metabolism regulating health and disease processes in the kidney. Context and significanceThe present study describes that endothelial cell SIRT3 protects against renal fibrosis by mitigating the metabolic reprogramming and associated mesenchymal transformations in diabetes. Accordingly, amelioration of endothelial cell SIRT3 and targeting the defective metabolism offer a potential therapeutic approach to treat or prevent the progression of diabetic kidney disease. Further investigation would delineate the mechanisms of SIRT3 up-regulation in diabetic kidney disease and discover small molecules to increase the expression of the SIRT3. In addition, this study describes the biology of SIRT3 in the health and disease processes of kidney endothelial cells.

cell biology↗

bioRxiv2020

Loss of endothelial glucocorticoid receptor accelerates diabetic nephropathy

Endothelial cells play a key role in the regulation of disease and other developmental processes. Defective regulation of endothelial cell homeostasis may cause mesenchymal activation of other endothelial cells by autocrine effects or of neighboring cell types by paracrine effects, and in both cases contribute to organ fibrosis. However, regulatory control of endothelial cell homeostasis, is not well studied. Diabetes induced renal fibrosis in endothelial GR knock out mice (GRfl/fl;Tie 1 Cre; GRECKO) but not in control mice (GRfl/fl); hypercholesterolemia further enhanced severe renal fibrosis in diabetic GRECKO; Apoe-/- (DKO) but not in diabetic littermates (GRfl/fl; Apoe-/-). The fibrogenic phenotype in the kidneys of diabetic GRECKO and diabetic DKO were associated with aberrant cytokine and chemokine reprogramming. Canonical Wnt signaling was identified as new target for the action of endothelial GR. Wnt inhibiton improved kidney fibrosis by mitigating endothelial-to-mesenchymal transition (EndMT) and epithelial-to-mesenchymal transitions (EMT). Similarly, activation of fatty acid oxidation also suppressed kidney fibrosis. Conditioned media from endothelial cells from diabetic GRECKO stimulated Wnt signaling-dependent epithelial-to-mesenchymal transition in tubular epithelial cells from diabetic controls. These data demonstrate that endothelial GR is an essential antifibrotic core molecule in diabetes.

cell biology↗