SMSwagatam Mukhopadhyay

Swagatam Mukhopadhyay

*This is an algorithmically generated author record based on publicly available information.

Swagatam Mukhopadhyay has a total of 14 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Sequence-Based Therapeutic Peptide Classification with Augmented Negative Sampling

    by Roman Ellerbrock, Alessio Valentini, Alexander Christian Paul, Swagatam Mukhopadhyay, Michael R. Perelshtein

    Therapeutic peptides offer high target specificity, low toxicity, and the ability to modulate protein-protein interactions, yet experimental functional characterization remains costly and slow. Computational prediction of therapeutic function directly from sequence could accelerate peptide screening and enable generative design pipelines, but requires reliable discrimination between therapeutic and non-therapeutic peptides. Existing multi-label predictors cover few functions, rely on limited datasets, and exhibit high \glspl{fpr}, limiting their practical utility. We present a lightweight CNN classifier trained on the most comprehensive therapeutic peptide database to date (54,655 peptides, 48 functional categories). A key contribution is a statistically motivated negative sampling strategy using Markov models to generate diverse synthetic decoys at multiple difficulty levels. When evaluated on this controlled decoy benchmark, the FRP is reduced from over 60% for previous models to 2.1% for our approach. Our fine-tuned five-model ensemble achieves 78.9% Micro F1 and 54.6% Macro F1 while requiring only amino acid sequences as inputs. Analysis using a sparse L1-constrained variant of our model shows that convolutional filters capture conserved functional motifs and statistically improbable non-therapeutic patterns, with downstream layers combining these signals, providing mechanistic evidence that the network learns biologically meaningful structure. In a generalization task on the TPpred-LE benchmark, our model achieves 55.3% Micro F1 and 38.6% Macro F1, comparable to TPpred-LE trained on its native dataset (57.9%/38.1%) while predicting four times more therapeutic functions with four times fewer parameters. Code and models will be made available at https://github.com/terra-quantum-public/tq-therapep-ai.

  • Open Access

    NucleoPhi: A Transferable Force Field for Modified Nucleic Acids Enabling Atomistic Simulation–Driven Oligonucleotide Design

    by Chungwen Liang, Swagatam Mukhopadhyay, David Pekker

    Accurate and transferable force fields are essential for reliably simulating both natural and chemically modified nucleic acids at atomistic resolution. However, there is no unified framework to consistently parameterize force fields across the wide range of sugar, base, and backbone or linker modifications commonly used in therapeutic oligonucleotides. Building on a generalized protocol, we develop force field parameters for the 2'-fluoro, 2'-O-methyl, and 2'FANA ribose modifications, as well as the pseudo-uridine nucleobase modification. These models were validated against a diverse set of experimentally determined structures and show good agreement with the observed conformational ensembles. We further demonstrate their utility in elucidating how chemical modifications modulate local structure, flexibility, and interaction energetics in nucleic acid hybrid duplexes, siRNA–Ago2 complexes, and the evasion of MC1 ribonuclease by pseudo-uridine. Overall, this work establishes a robust, generalizable pipeline for developing force fields for modified nucleic acids, enabling predictive atomistic simulations reliably.


  • Toxicity of Antisense Oligonucleotides is Determined by the Synergistic Interplay of Chemical Modifications and Nucleotide Sequences, Not by Either Factor Alone

    by Jaspreet Bhamra, Mahati Krishna, George Samaan, Sankha Pattanayak, Swagatam Mukhopadhyay

    Antisense oligonucleotides (ASOs) offer a promising therapeutic approach for precise RNA‐level gene modulation. Despite advancements in chemical modifications to enhance stability and pharmacokinetics, ASOs still face significant challenges, including liver, immunological, renal, and neurological toxicities, potentially leading to high preclinical failure rates. Current oligonucleotide‐based drug optimization strategies to overcome such issues often rely on applying a few commonly used chemical architectures (patterns of linker, sugar, or base modifications), which are conventional in the field, or engaging in expensive and time‐consuming trial‐and‐error screening processes involving both sequence changes and positional chemical modifications. These traditional approaches treat nucleotide sequences (“sequence”) and sugar/linkage modification chemistries (“chemistry”) as independent contributors to toxicity. However, ample evidence in the literature shows that even minor changes in either sequence or chemical modifications can drastically impact toxicity, suggesting an inseparable synergistic relationship between sequence and chemistry. In support of this sequence‐chemistry collusion thesis, this manuscript presents a survey of the systemic toxicity potential of several chemically modified gapmer ASOs by investigating the impact of modifying sugar and backbone chemistries on ASO‐induced hepatotoxicity, nephrotoxicity, and immune/inflammatory responses. The data unequivocally demonstrate that ASO toxicity is strongly influenced by the interplay between nucleotide sequence, chemical modifications, and the specific position context of those modifications, highlighting the critical need to rationally engineer the optimal sequence and chemical composition to develop safe and active ASO drug candidates instead of discovering suboptimal ASOs through trial‐and‐error screening campaigns.


  • Open Access

    A Dose-Response Model for Accurate Detection and Quantification of Transcriptome-Wide Gene Knockdown for Oligonucleotide-Based Medicines

    by David Pekker, Steven Kuntz, Monica McArthur, Tim Nicholson-Shaw, Sara Yanke, Swagatam Mukhopadhyay

    Synthetic antisense oligonucleotides and siRNAs are a class of Oligonucleotide-Based Medicines (OBMs) that can hybridize with pre-mRNA and mRNA, recruit a mechanism-of-action specific enzymatic complex, and knockdown target gene expression. This class of molecules provides an excellent substrate for designing precision gene-modulatory therapeutics; however, quantifying on- and off-target dose response as measured by next-generation sequencing for this class of therapeutics has remained under-powered and ambiguous. Often in silico predictions of off-targets (ranked by edit tolerance) are used as putative off-target analysis in ASO and siRNA drug design. We construct a simple, effective theory of transcriptional dynamics and enzymatic activity in order to describe the transcriptome-wide response to these oligonucleotides. We establish rigorous quantification methods of off-target analysis in oligonucleotide drug design. We also extend the DESeq work [1, 2] of Negative Binomial noise in gene expression measurements to describe noise, including outliers, in OBM-dose response NGS experiments. We demonstrate the performance of our model on both synthetic and experimental Digital Gene Expression (DGE) data of dose response in ASO-treated cells. We present our analysis package, DoReSeq , as a freely available resource for the community. We hope this will elevate the standards of off-target analysis for such an important class of precision therapeutics.