A high-resolution layer-by-layer interpretability analysis of 8 models from the ESM2 and AMPLIFY families on 22 concepts from human proteome annotations found that these models encode concepts of increasing levels of complexity along their depth: basic physicochemical properties and linear motifs are best captured by early-layer embeddings, secondary structure from subsequent layers, and domain-level semantics from middle layers.
Abstract
Protein language models have been increasingly successful on tasks ranging from fitness prediction to functional design, yet what biological knowledge they acquire and where it is encoded within their internal representations remain underexplored. Through a high-resolution layer-by-layer interpretability analysis of 8 models from the ESM2 and AMPLIFY families on 22 concepts from human proteome annotations, we found that these models encode concepts of increasing levels of complexity along their depth: basic physicochemical properties and linear motifs are best captured by early-layer embeddings, secondary structure from subsequent layers, and domain-level semantics from middle layers. Principal component projections of these embeddings showed that they separate biologically meaningful protein groupings, and molecular-biology-inspired interventions demonstrated that pLM embeddings can discriminate phosphomimic-active from inactive mutants. Perhaps surprisingly, we observed that pretraining data and compute had a greater impact on the linear emergence of biological concepts than scaling up parameters. By revealing where biological knowledge is captured in pLMs and which choices shape its emergence, our work offers insights to develop more robust, biologically grounded protein language models.
Protein language models (PLMs) have transferred the latest advances from natural language processing to computational biology. These models, trained on large corpora of protein sequence data, are widely used to translate amino acid sequences into latent-space embeddings, ready for use in diverse downstream tasks (DTs). By a common consensus, embeddings from the model's last layer are used, and the model's internal behavior remains poorly understood. We analyzed 13 PLMs across 15 DTs from 11 datasets to investigate the informativeness of embeddings created in intermediate PLM layers. We trained probe models on embeddings from each layer, compared their performance, and computed characteristics of the latent spaces they span to estimate the information they contain, and found that the last layers of PLMs rarely contained embeddings that led to the best results on downstream tasks. Furthermore, we identified a connection between DTs and the distribution across PLMs'layers of the relevant information to predict that task. For example, similarity between the pre-training objective and the objective of predicting properties of individual residues leads to a steady increase in understanding of such tasks across the layers of PLMs. On the other hand, for whole-protein tasks, we observe that the dataset, rather than the task itself, defines PLMs'ability to perform well on a DT. Embeddings from shallow layers of PLMs perform better for datasets that contain deep mutational scan (DMS) data, while datasets containing diverse natural proteins find most useful embeddings in the models'deeper layers. Additionally, we discover that the performance of PLMs drops significantly when tasks are introduced for artificial proteins.
R. Joeres, Ilya S. Senatorov, A. Kolchina et al.· 0 citations
The utility of HA sites for suggesting candidate binding sites and the biological interpretability of PLM representations is explored, demonstrating the biological interpretability of PLM representations and offers a valuable method to prioritize functionally relevant protein residues for targeted biomedical research.
Sophia J. Pribus, Russ B. Altman, Gowri Nayar· bioRxiv· 0 citations
Protein language models (pLMs) encode information about protein sequences which enable downstream tasks such as structure prediction, but their internal representations are not well understood. Sparse autoencoders (SAEs) provide a promising tool to disentangle latent pLM representations into interpretable features, but existing annotation pipelines largely rely on protein-level annotations derived from database labels and LLM annotations of top activating sequences. Such annotations can overlook the localized residue-level and geometric patterns encoded by sparse features. We introduce an automated and scalable method for interpreting SAE features in ESM-2 by using geometrically inspired features of the protein $\text{C}_\alpha$ backbone. Across ESM-2 8M layers, an FDR-controlled discovery analysis shows that local geometry is significantly associated with many SAE features, with varying levels of predictive strength, expanding coverage beyond database and sequence-based methods. In particular, geometry can distinguish SAE features sharing the same database annotation, revealing substructure within known biological labels. A significant portion of SAE features activate on unannotated metagenomic protein sequences enabling us to use our SAE annotations to better understand these sequences. In addition, ablation experiments at the level of contact prediction show that removing found geometric features shifts ESM-2's predicted contact maps in the direction of the descriptor. This provides a robust method of annotating proteins activated within SAE neurons at a residue level, providing a bridge between mechanistic interpretability and structural biology.
S. Setlur, Djordje Mihajlovic, Darrick Lee· 0 citations
Recently, there has been a growing adoption of protein language models (PLMs) in biomedical science. Their embeddings provide a rich numerical representation of protein sequences which achieve state-of-the-art performance on several downstream tasks including protein fitness prediction. However, PLM embeddings are not directly interpretable and, thereby, it remains unclear what features they encode. To gain insight into which biochemical properties of the protein are driving the prediction, we leverage an orthogonal projection technique that removes linear effects of known tabular features from embeddings and extend it to high-order and interaction effects. In this way, we remove the effects of interpretable biochemical features from PLM embeddings. In an ablation study, we show that this leads to a decrease in performance for a downstream classifier trained only on the embeddings to predict protein fitness. In an additional evaluation, we find that these biochemical features explain a substantial part of the variance in the predictions of this classifier. Hence, we can show that PLM embeddings encode patterns correlated with biochemical properties and quantify their contribution to predicting protein fitness. This computationally efficient approach is not limited to the features or embeddings considered here and is readily transferable to problem settings beyond protein fitness prediction.
Paulo Yanez Sarmiento, Pia Francesca Rissom, Manuel Pfeuffer et al.· 0 citations
Applications to thioredoxins, visual opsins, and Tara Oceans environmental diatom cold-shock proteins show that PLMView can move from interpretable residue-level determinants in well-studied protein families to large-scale environmental functional discovery, linking molecular specialization to ecological distribution and transcriptional deployment across the global ocean.
This work argues that explicitly modeling at the functional motif level provides both mechanistic insight into sequence-function relationships and interpretable control over protein generation, an important step toward compositional design of novel protein functions.
Boon How Low, W. Goh, Boyang Li et al.· Proceedings of the 32nd ACM...· 0 citations
