Protein Representation Learning

Unlike the predictable scaling laws in natural language processing and computer vision, protein language models (PLMs) scale poorly: for many tasks, models within the same family plateau or even decrease in performance, with mid-sized models often outperforming the largest in the family. We introduce Reverse Distillation, a principled framework that decomposes large PLM representations into orthogonal subspaces guided by smaller models of the same family. The resulting embeddings have a nested, Matryoshka-style structure: the first k dimensions of a larger model's embedding are exactly the representation from the smaller model. This ensures that larger reverse-distilled models consistently outperform smaller ones. A motivating intuition is that smaller models, constrained by capacity, preferentially encode broadly-shared protein features. Reverse distillation isolates these shared features and orthogonally extracts additional contributions from larger models, preventing interference between the two. On ProteinGym benchmarks, reverse-distilled ESM-2 variants outperform their respective baselines at the same embedding dimensionality, with the reverse-distilled 15 billion parameter model achieving the strongest performance. Our framework is generalizable to any model family where scaling challenges persist. Code and trained models are available at https://github.com/rohitsinghlab/plm_reverse_distillation.

Protein language models often take into consideration the alignment between a protein sequence and its textual description. However, they do not take structural information into consideration. Traditional methods treat sequence and structure separately, limiting the ability to exploit the alignment between the structure and protein sequence embeddings. In this paper, we introduce a sequence structure contrastive alignment framework, which learns a shared embedding space where proteins are represented consistently across modalities. By training on large-scale pairs of sequences and experimentally resolved or predicted structures, the model maximizes agreement between matched sequence structure pairs while pushing apart unrelated pairs. This alignment enables cross-modal retrieval (e.g., finding structural neighbors given a sequence), improves downstream prediction tasks such as function annotation and stability estimation, and provides interpretable links between sequence variation and structural organization. Our results demonstrate that contrastive learning can serve as a powerful bridge between protein sequences and structures, offering a unified representation for understanding and engineering proteins.

Generative models have recently advanced $\textit{de novo}$ protein design by learning the statistical regularities of natural structures. However, current approaches face three key limitations: (1) Existing methods cannot jointly learn protein geometry and design tasks, where pretraining can be a solution; (2) Current pretraining methods mostly rely on local, non-rigid atomic representations for property prediction downstream tasks, limiting global geometric understanding for protein generation tasks; and (3) Existing approaches have yet to effectively model the rich dynamic and conformational information of protein structures. To overcome these issues, we introduce $\textbf{RigidSSL}$ ($\textit{Rigidity-Aware Self-Supervised Learning}$), a geometric pretraining framework that front-loads geometry learning prior to generative finetuning. Phase I (RigidSSL-Perturb) learns geometric priors from 432K structures from the AlphaFold Protein Structure Database with simulated perturbations. Phase II (RigidSSL-MD) refines these representations on 1.3K molecular dynamics trajectories to capture physically realistic transitions. Underpinning both phases is a bi-directional, rigidity-aware flow matching objective that jointly optimizes translational and rotational dynamics to maximize mutual information between conformations. Empirically, RigidSSL variants improve designability by up to 43% while enhancing novelty and diversity in unconditional generation. Furthermore, RigidSSL-Perturb improves the success rate by 5.8% in zero-shot motif scaffolding and RigidSSL-MD captures more biophysically realistic conformational ensembles in G protein-coupled receptor modeling.

Protein sequences are abundant in repeating segments, both as exact copies and as approximate segments with mutations. These repeats are important for protein structure and function, motivating decades of algorithmic work on repeat identification. Recent work has shown that protein language models (PLMs) identify repeats, by examining their behavior in masked-token prediction. To elucidate their internal mechanisms, we investigate how PLMs detect both exact and approximate repeats. We find that the mechanism for approximate repeats functionally subsumes that of exact repeats. We then characterize this mechanism, revealing two main stages: PLMs first build feature representations using both general positional attention heads and biologically specialized components, such as neurons that encode amino-acid similarity. Then, induction heads attend to aligned tokens across repeated segments, promoting the correct answer. Our results reveal how PLMs solve this biological task by combining language-based pattern matching with specialized biological knowledge, thereby establishing a basis for studying more complex evolutionary processes in PLMs.

Protein language models (PLMs) have become widely adopted as general-purpose models, demonstrating strong performance in protein engineering and de novo design. Like large language models (LLMs), they are typically trained as deep transformers with next-token or masked-token prediction objectives on massive sequence corpora and are scaled by increasing model depth. Recent work on autoregressive LLMs has identified the Curse of Depth: many later layers contribute little to the final output predictions. These findings naturally raise the question of whether a similar depth inefficiency also appears in PLMs, where many widely used models are not autoregressive, and some are multimodal, accepting both protein sequence and structure as input. In this work, we present a depth analysis of seven popular PLM families across model scales, spanning autoregressive, masked, and diffusion objectives, and quantify how layer contributions evolve with depth using a unified set of probing-, perturbation-, and downstream-evaluation measurements. Across models, we observe consistent depth-dependent patterns that extend prior findings on LLMs: a large fraction of task-relevant computation is concentrated in a subset of layers, while the remaining layers mainly provide incremental refinement of the final prediction. These trends persist beyond sequence-only settings and also appear in multimodal PLMs. Taken together, our results suggest that depth inefficiency is a common feature of modern PLMs, motivating future work on more depth-efficient architectures and training methods.

Modern Protein Language Models (PLMs) apply transformer-based model architectures from natural language processing to biological sequences, predicting a variety of protein functions and properties. However, protein language has key differences from natural language, such as a rich functional space despite a vocabulary of only 20 amino acids. These differences motivate research into how transformer-based architectures operate differently in the protein domain and how we can better leverage PLMs to solve protein-related tasks. In this work, we begin by directly comparing how the distribution of information stored across layers of attention heads differs between the protein and natural language domain. Furthermore, we adapt a simple early-exit technique-originally used in the natural language domain to improve efficiency at the cost of performance-to achieve both increased accuracy and substantial efficiency gains in protein non-structural property prediction by allowing the model to automatically select protein representations from the intermediate layers of the PLMs for the specific task and protein at hand. We achieve performance gains ranging from 0.4 to 7.01 percentage points while simultaneously improving efficiency by over 10 percent across models and non-structural prediction tasks. Our work opens up an area of research directly comparing how language models change behavior when moved into the protein domain and advances language modeling in biological domains.

Effective representations of protein sequences are widely recognized as a cornerstone of machine learning-based protein design. Yet, protein bioengineering poses unique challenges for sequence representation, as experimental datasets typically feature few mutations, which are either sparsely distributed across the entire sequence or densely concentrated within localized regions. This limits the ability of sequence-level representations to extract functionally meaningful signals. In addition, comprehensive comparative studies remain scarce, despite their crucial role in clarifying which representations best encode relevant information and ultimately support superior predictive performance. In this study, we systematically evaluate multiple ProtBERT and ESM2 embedding variants as sequence representations, using the adeno-associated virus capsid as a case study and prototypical example of bioengineering, where functional optimization is targeted through highly localized sequence variation within an otherwise large protein. Our results reveal that, prior to fine-tuning, amino acid-level embeddings outperform sequence-level representations in supervised predictive tasks, whereas the latter tend to be more effective in unsupervised settings. However, optimal performance is only achieved when embeddings are fine-tuned with task-specific labels, with sequence-level representations providing the best performance. Moreover, our findings indicate that the extent of sequence variation required to produce notable shifts in sequence representations exceeds what is typically explored in bioengineering studies, showing the need for fine-tuning in datasets characterized by sparse or highly localized mutations.

Motivation: Pre-trained protein language models (PLMs) such as ESM-2 have become the default representation for biological sequence tasks, but they are computationally heavy and require GPUs both for embedding and for fine-tuning. Whether they are actually necessary for sequence classification, as opposed to structure prediction, is rarely tested against strong, principled, lightweight alternatives. This question has direct practical stakes for large-scale genomic surveillance, where embedding millions of sequences on commodity hardware is a recurring bottleneck.\ Results: We introduce Murmur2Vec, an alignment-free, training-free embedding that aggregates $k$-mer counts into a small hash table via the deterministic MurmurHash function, and we cast it as a randomized sketch of the classical $k$-mer spectrum kernel. We provide a complete theoretical treatment: closed-form bias/variance of the inner product, an unbiased signed variant with a Johnson--Lindenstrauss-type concentration bound, an excess-risk bound for downstream linear classifiers that makes the bias--variance trade-off in the hash-table size explicit, and an implicit-regularization mechanism by which collisions damage frequent non-discriminative $k$-mers more than rare lineage-defining ones. Across four classification tasks, SARS-CoV-2 spike lineage (22 classes), HIV-1 Env subtype (8 classes), and two protein-family benchmarks (8 and 6 classes), Murmur2Vec matches a LoRA-fine-tuned 650M-parameter ESM-2 model on the two tasks for which LoRA fine-tuning was run to convergence (SARS-CoV-2 and HIV-1) and ties frozen ESM-2 on the two protein-family tasks, and it outperforms the fine-tuned model on the hardest task (SARS-CoV-2 lineage: $0.854$ vs.\ $0.807$ accuracy; macro-F1 $0.684$ vs.\ $0.401$).

Modeling protein sequences as a language has made language models a powerful tool in computational biology, yet the language itself remains poorly understood. A key step toward understanding it is identifying its constituent units. In natural languages, morphemes can occur in multiple forms; similarly, in proteins, mutations can give rise to variations of a unit that persist through evolution, forming families of related units. We introduce PUMA (Protein Units via Mutation-Aware Merging), an algorithm that learns protein units from sequence and explores their mutational variants using substitution matrices, forming a genealogy of unit families. Our results show that mutations remaining within a PUMA family are more often benign than the substitution matrix alone predicts, and that PUMA genealogy improves molecular function representations compared to treating units independently. A case study of a unit family demonstrates relatedness beyond homology. PUMA achieves competitive performance on downstream tasks when used as a protein language model tokenizer. Moreover, collapsing units into families results in a smaller embedding table and faster training. Together, these results support PUMA as a biologically grounded protein vocabulary that organizes protein units into plausible families of mutational variants. The source code is available at https://github.com/boun-tabi-lifelu/PUMA.