arXiv subjects

Melih Yilmaz

Publications and source records attributed to Melih Yilmaz.

3 recordsLinked to original sources

arXiv2026

Conditional generation of antibody sequences with classifier-guided germline-absorbing discrete diffusion

Antibody therapeutics are among the most successful modern medicines, yet computationally designing antibodies with desirable binding and developability properties remains challenging. While protein language models (pLMs) have emerged as powerful tools for antibody sequence design, existing approaches largely suffer from two key limitations: they predominantly memorize germline sequences rather than modeling biologically meaningful somatic variation, and they offer limited support for flexible classifier-guided conditional generation. We address these challenges through two primary contributions. First, we demonstrate that discrete diffusion fine-tuning achieves strong language modeling performance on antibody sequences while allowing for generation conditioned on any off-the-shelf classifier. Second, we introduce germline absorbing diffusion, a novel modification of the discrete diffusion noise process in which the germline sequence - rather than a masked sequence - serves as the absorbing state. This biologically motivated inductive bias restricts the model to learning the trajectory from germline to observed sequence, effectively excluding genetic variation and V(D)J recombination statistics from the learned distribution and dramatically mitigating germline bias. We show that germline diffusion improves non-germline residue prediction accuracy from 26 percent to 46 percent, approaching the theoretical upper bound set by true biological variability. We then demonstrate the utility of our germline diffusion model on the conditional generation tasks of sampling antibodies with improved hydrophobicity and predicted binding affinity. On both tasks our model shows an improved tradeoff between class adherence and sample quality, significantly outperforming EvoProtGrad, a popular strategy to sample from pLMs with gradient-based discrete Markov Chain Monte Carlo.

cs.LG↗

arXiv2025

Pi-SAGE: Permutation-invariant surface-aware graph encoder for binding affinity prediction

Protein surface fingerprint encodes chemical and geometric features that govern protein-protein interactions and can be used to predict changes in binding affinity between two protein complexes. Current state-of-the-art models for predicting binding affinity change, such as GearBind, are all-atom based geometric models derived from protein structures. Although surface properties can be implicitly learned from the protein structure, we hypothesize that explicit knowledge of protein surfaces can improve a structure-based model's ability to predict changes in binding affinity. To this end, we introduce Pi-SAGE, a novel Permutation-Invariant Surface-Aware Graph Encoder. We first train Pi-SAGE to create a protein surface codebook directly from the structure and assign a token for each surface-exposed residue. Next, we augment the node features of the GearBind model with surface features from domain-adapted Pi-SAGE to predict binding affinity change on the SKEMPI dataset. We show that explicitly incorporating local, context-aware chemical properties of residues enhances the predictive power of all-atom graph neural networks in modeling binding affinity changes between wild-type and mutant proteins.

q-bio.BM↗

arXiv2025

Foundation model for mass spectrometry proteomics

Mass spectrometry is the dominant technology in the field of proteomics, enabling high-throughput analysis of the protein content of complex biological samples. Due to the complexity of the instrumentation and resulting data, sophisticated computational methods are required for the processing and interpretation of acquired mass spectra. Machine learning has shown great promise to improve the analysis of mass spectrometry data, with numerous purpose-built methods for improving specific steps in the data acquisition and analysis pipeline reaching widespread adoption. Here, we propose unifying various spectrum prediction tasks under a single foundation model for mass spectra. To this end, we pre-train a spectrum encoder using de novo sequencing as a pre-training task. We then show that using these pre-trained spectrum representations improves our performance on the four downstream tasks of spectrum quality prediction, chimericity prediction, phosphorylation prediction, and glycosylation status prediction. Finally, we perform multi-task fine-tuning and find that this approach improves the performance on each task individually. Overall, our work demonstrates that a foundation model for tandem mass spectrometry proteomics trained on de novo sequencing learns generalizable representations of spectra, improves performance on downstream tasks where training data is limited, and can ultimately enhance data acquisition and analysis in proteomics experiments.

cs.LG↗