Tree visualizations of protein sequence embedding space enable improved functional clustering of diverse protein superfamilies.

Protein language models, trained on millions of biologically observed sequences, generate feature-rich numerical representations of protein sequences. These representations, called sequence embeddings, can infer structure-functional properties, despite protein language models being trained on primary sequence alone. While sequence embeddings have been applied toward tasks such as structure and function prediction, applications toward alignment-free sequence classification have been hindered by the lack of studies to derive, quantify and evaluate relationships between protein sequence embeddings. Here, we develop workflows and visualization methods for the classification of protein families using sequence embedding derived from protein language models. A benchmark of manifold visualization methods reveals that Neighbor Joining (NJ) embedding trees are highly effective in capturing global structure while achieving similar performance in capturing local structure compared with popular dimensionality reduction techniques such as t-SNE and UMAP. The statistical significance of hierarchical clusters on a tree is evaluated by resampling embeddings using a variational autoencoder (VAE). We demonstrate the application of our methods in the classification of two well-studied enzyme superfamilies, phosphatases and protein kinases. Our embedding-based classifications remain consistent with and extend upon previously published sequence alignment-based classifications. We also propose a new hierarchical classification for the S-Adenosyl-L-Methionine (SAM) enzyme superfamily which has been difficult to classify using traditional alignment-based approaches. Beyond applications in sequence classification, our results further suggest NJ trees are a promising general method for visualizing high-dimensional data sets.

Using explainable machine learning to uncover the kinase-substrate interaction landscape.

MOTIVATION: Phosphorylation, a post-translational modification regulated by protein kinase enzymes, plays an essential role in almost all cellular processes. Understanding how each of the nearly 500 human protein kinases selectively phosphorylates their substrates is a foundational challenge in bioinformatics and cell signaling. Although deep learning models have been a popular means to predict kinase-substrate relationships, existing models often lack interpretability and are trained on datasets skewed toward a subset of well-studied kinases. RESULTS: Here we leverage recent peptide library datasets generated to determine substrate specificity profiles of 300 serine/threonine kinases to develop an explainable Transformer model for kinase-peptide interaction prediction. The model, trained solely on primary sequences, achieved state-of-the-art performance. Its unique multitask learning paradigm built within the model enables predictions on virtually any kinase-peptide pair, including predictions on 139 kinases not used in peptide library screens. Furthermore, we employed explainable machine learning methods to elucidate the model's inner workings. Through analysis of learned embeddings at different training stages, we demonstrate that the model employs a unique strategy of substrate prediction considering both substrate motif patterns and kinase evolutionary features. SHapley Additive exPlanation (SHAP) analysis reveals key specificity determining residues in the peptide sequence. Finally, we provide a web interface for predicting kinase-substrate associations for user-defined sequences and a resource for visualizing the learned kinase-substrate associations. AVAILABILITY AND IMPLEMENTATION: All code and data are available at https://github.com/esbgkannan/Phosformer-ST. Web server is available at https://phosformer.netlify.app.

Cancer-associated mutations in protein kinase C theta are loss-of-function.

The Ca2+-independent, but diacylglycerol-regulated, novel protein kinase C (PKC) theta (θ) is highly expressed in hematopoietic cells where it participates in immune signaling and platelet function. Mounting evidence suggests that PKCθ may be involved in cancer, particularly blood cancers, breast cancer, and gastrointestinal stromal tumors, yet how to target this kinase (as an oncogene or as a tumor suppressor) has not been established. Here, we examine the effect of four cancer-associated mutations, R145H/C in the autoinhibitory pseudosubstrate, E161K in the regulatory C1A domain, and R635W in the regulatory C-terminal tail, on the cellular activity and stability of PKCθ. Live-cell imaging studies using the genetically-encoded fluorescence resonance energy transfer-based reporter for PKC activity, C kinase activity reporter 2 (CKAR2), revealed that the pseudosubstrate and C1A domain mutations impaired autoinhibition to increase basal signaling. This impaired autoinhibition resulted in decreased stability of the protein, consistent with the well-characterized behavior of Ca2+-regulated PKC isozymes wherein mutations that impair autoinhibition are paradoxically loss-of-function because the mutant protein is degraded. In marked contrast, the C-terminal tail mutation resulted in enhanced autoinhibition and enhanced stability. Thus, the examined mutations were loss-of-function by different mechanisms: mutations that impaired autoinhibition promoted the degradation of PKC, and those that enhanced autoinhibition stabilized an inactive PKC. Supporting a general loss-of-function of PKCθ in cancer, bioinformatics analysis revealed that protein levels of PKCθ are reduced in diverse cancers, including lung, renal, head and neck, and pancreatic. Our results reveal that PKCθ function is lost in cancer.

Phosformer: an explainable transformer model for protein kinase-specific phosphorylation predictions.

MOTIVATION: The human genome encodes over 500 distinct protein kinases which regulate nearly all cellular processes by the specific phosphorylation of protein substrates. While advances in mass spectrometry and proteomics studies have identified thousands of phosphorylation sites across species, information on the specific kinases that phosphorylate these sites is currently lacking for the vast majority of phosphosites. Recently, there has been a major focus on the development of computational models for predicting kinase-substrate associations. However, most current models only allow predictions on a subset of well-studied kinases. Furthermore, the utilization of hand-curated features and imbalances in training and testing datasets pose unique challenges in the development of accurate predictive models for kinase-specific phosphorylation prediction. Motivated by the recent development of universal protein language models which automatically generate context-aware features from primary sequence information, we sought to develop a unified framework for kinase-specific phosphosite prediction, allowing for greater investigative utility and enabling substrate predictions at the whole kinome level. RESULTS: We present a deep learning model for kinase-specific phosphosite prediction, termed Phosformer, which predicts the probability of phosphorylation given an arbitrary pair of unaligned kinase and substrate peptide sequences. We demonstrate that Phosformer implicitly learns evolutionary and functional features during training, removing the need for feature curation and engineering. Further analyses reveal that Phosformer also learns substrate specificity motifs and is able to distinguish between functionally distinct kinase families. Benchmarks indicate that Phosformer exhibits significant improvements compared to the state-of-the-art models, while also presenting a more generalized, unified, and interpretable predictive framework. AVAILABILITY AND IMPLEMENTATION: Code and data are available at https://github.com/esbgkannan/phosformer. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Modularity of the hydrophobic core and evolution of functional diversity in fold A glycosyltransferases.

Hydrophobic cores are fundamental structural properties of proteins typically associated with protein folding and stability; however, how the hydrophobic core shapes protein evolution and function is poorly understood. Here, we investigated the role of conserved hydrophobic cores in fold-A glycosyltransferases (GT-As), a large superfamily of enzymes that catalyze formation of glycosidic linkages between diverse donor and acceptor substrates through distinct catalytic mechanisms (inverting versus retaining). Using hidden Markov models and protein structural alignments, we identify similarities in the phosphate-binding cassette (PBC) of GT-As and unrelated nucleotide-binding proteins, such as UDP-sugar pyrophosphorylases. We demonstrate that GT-As have diverged from other nucleotide-binding proteins through structural elaboration of the PBC and its unique hydrophobic tethering to the F-helix, which harbors the catalytic base (xED-Asp). While the hydrophobic tethering is conserved across diverse GT-A fold enzymes, some families, such as B3GNT2, display variations in tethering interactions and core packing. We evaluated the structural and functional impact of these core variations through experimental mutational analysis and molecular dynamics simulations and find that some of the core mutations (T336I in B3GNT2) increase catalytic efficiency by modulating the conformational occupancy of the catalytic base between "D-in" and acceptor-accessible "D-out" conformation. Taken together, our studies support a model of evolution in which the GT-A core evolved progressively through elaboration upon an ancient PBC found in diverse nucleotide-binding proteins, and malleability of this core provided the structural framework for evolving new catalytic and substrate-binding functions in extant GT-A fold enzymes.

KinOrtho: a method for mapping human kinase orthologs across the tree of life and illuminating understudied kinases.

BACKGROUND: Protein kinases are among the largest druggable family of signaling proteins, involved in various human diseases, including cancers and neurodegenerative disorders. Despite their clinical relevance, nearly 30% of the 545 human protein kinases remain highly understudied. Comparative genomics is a powerful approach for predicting and investigating the functions of understudied kinases. However, an incomplete knowledge of kinase orthologs across fully sequenced kinomes severely limits the application of comparative genomics approaches for illuminating understudied kinases. Here, we introduce KinOrtho, a query- and graph-based orthology inference method that combines full-length and domain-based approaches to map one-to-one kinase orthologs across 17 thousand species. RESULTS: Using multiple metrics, we show that KinOrtho performed better than existing methods in identifying kinase orthologs across evolutionarily divergent species and eliminated potential false positives by flagging sequences without a proper kinase domain for further evaluation. We demonstrate the advantage of using domain-based approaches for identifying domain fusion events, highlighting a case between an understudied serine/threonine kinase TAOK1 and a metabolic kinase PIK3C2A with high co-expression in human cells. We also identify evolutionary fission events involving the understudied OBSCN kinase domains, further highlighting the value of domain-based orthology inference approaches. Using KinOrtho-defined orthologs, Gene Ontology annotations, and machine learning, we propose putative biological functions of several understudied kinases, including the role of TP53RK in cell cycle checkpoint(s), the involvement of TSSK3 and TSSK6 in acrosomal vesicle localization, and potential functions for the ULK4 pseudokinase in neuronal development. CONCLUSIONS: In sum, KinOrtho presents a novel query-based tool to identify one-to-one orthologous relationships across thousands of proteomes that can be applied to any protein family of interest. We exploit KinOrtho here to identify kinase orthologs and show that its well-curated kinome ortholog set can serve as a valuable resource for illuminating understudied kinases, and the KinOrtho framework can be extended to any protein-family of interest.

Informatic challenges and advances in illuminating the druggable proteome.

The understudied members of the druggable proteomes offer promising prospects for drug discovery efforts. While large-scale initiatives have generated valuable functional information on understudied members of the druggable gene families, translating this information into actionable knowledge for drug discovery requires specialized informatics tools and resources. Here, we review the unique informatics challenges and advances in annotating understudied members of the druggable proteome. We demonstrate the application of statistical evolutionary inference tools, knowledge graph mining approaches, and protein language models in illuminating understudied protein kinases, pseudokinases, and ion channels.

Predicting protein and pathway associations for understudied dark kinases using pattern-constrained knowledge graph embedding.

The 534 protein kinases encoded in the human genome constitute a large druggable class of proteins that include both well-studied and understudied "dark" members. Accurate prediction of dark kinase functions is a major bioinformatics challenge. Here, we employ a graph mining approach that uses the evolutionary and functional context encoded in knowledge graphs (KGs) to predict protein and pathway associations for understudied kinases. We propose a new scalable graph embedding approach, RegPattern2Vec, which employs regular pattern constrained random walks to sample diverse aspects of node context within a KG flexibly. RegPattern2Vec learns functional representations of kinases, interacting partners, post-translational modifications, pathways, cellular localization, and chemical interactions from a kinase-centric KG that integrates and conceptualizes data from curated heterogeneous data resources. By contextualizing information relevant to prediction, RegPattern2Vec improves accuracy and efficiency in comparison to other random walk-based graph embedding approaches. We show that the predictions produced by our model overlap with pathway enrichment data produced using experimentally validated Protein-Protein Interaction (PPI) data from both publicly available databases and experimental datasets not used in training. Our model also has the advantage of using the collected random walks as biological context to interpret the predicted protein-pathway associations. We provide high-confidence pathway predictions for 34 dark kinases and present three case studies in which analysis of meta-paths associated with the prediction enables biological interpretation. Overall, RegPattern2Vec efficiently samples multiple node types for link prediction on biological knowledge graphs and the predicted associations between understudied kinases, pseudokinases, and known pathways serve as a conceptual starting point for hypothesis generation and testing.

Dark kinase annotation, mining, and visualization using the Protein Kinase Ontology.

The Protein Kinase Ontology (ProKinO) is an integrated knowledge graph that conceptualizes the complex relationships among protein kinase sequence, structure, function, and disease in a human and machine-readable format. In this study, we have significantly expanded ProKinO by incorporating additional data on expression patterns and drug interactions. Furthermore, we have developed a completely new browser from the ground up to render the knowledge graph visible and interactive on the web. We have enriched ProKinO with new classes and relationships that capture information on kinase ligand binding sites, expression patterns, and functional features. These additions extend ProKinO's capabilities as a discovery tool, enabling it to uncover novel insights about understudied members of the protein kinase family. We next demonstrate the application of ProKinO. Specifically, through graph mining and aggregate SPARQL queries, we identify the p21-activated protein kinase 5 (PAK5) as one of the most frequently mutated dark kinases in human cancers with abnormal expression in multiple cancers, including a previously unappreciated role in acute myeloid leukemia. We have identified recurrent oncogenic mutations in the PAK5 activation loop predicted to alter substrate binding and phosphorylation. Additionally, we have identified common ligand/drug binding residues in PAK family kinases, underscoring ProKinO's potential application in drug discovery. The updated ontology browser and the addition of a web component, ProtVista, which enables interactive mining of kinase sequence annotations in 3D structures and Alphafold models, provide a valuable resource for the signaling community. The updated ProKinO database is accessible at https://prokino.uga.edu.

Mechanistic and evolutionary insights into isoform-specific 'supercharging' in DCLK family kinases.

Catalytic signaling outputs of protein kinases are dynamically regulated by an array of structural mechanisms, including allosteric interactions mediated by intrinsically disordered segments flanking the conserved catalytic domain. The Doublecortin Like Kinases (DCLKs) are a family of microtubule-associated proteins characterized by a flexible C-terminal autoregulatory 'tail' segment that varies in length across the various human DCLK isoforms. However, the mechanism whereby these isoform-specific variations contribute to unique modes of autoregulation is not well understood. Here, we employ a combination of statistical sequence analysis, molecular dynamics simulations and in vitro mutational analysis to define hallmarks of DCLK family evolutionary divergence, including analysis of splice variants within the DCLK1 sub-family, which arise through alternative codon usage and serve to 'supercharge' the inhibitory potential of the DCLK1 C-tail. We identify co-conserved motifs that readily distinguish DCLKs from all other Calcium Calmodulin Kinases (CAMKs), and a 'Swiss-army' assembly of distinct motifs that tether the C-terminal tail to conserved ATP and substrate-binding regions of the catalytic domain to generate a scaffold for auto-regulation through C-tail dynamics. Consistently, deletions and mutations that alter C-terminal tail length or interfere with co-conserved interactions within the catalytic domain alter intrinsic protein stability, nucleotide/inhibitor-binding, and catalytic activity, suggesting isoform-specific regulation of activity through alternative splicing. Our studies provide a detailed framework for investigating kinome-wide regulation of catalytic output through cis-regulatory events mediated by intrinsically disordered segments, opening new avenues for the design of mechanistically-divergent DCLK1 modulators, stabilizers or degraders.

Mechanistic and evolutionary insights into isoform-specific 'supercharging' in DCLK family kinases.

Catalytic signaling outputs of protein kinases are dynamically regulated by an array of structural mechanisms, including allosteric interactions mediated by intrinsically disordered segments flanking the conserved catalytic domain. The doublecortin-like kinases (DCLKs) are a family of microtubule-associated proteins characterized by a flexible C-terminal autoregulatory 'tail' segment that varies in length across the various human DCLK isoforms. However, the mechanism whereby these isoform-specific variations contribute to unique modes of autoregulation is not well understood. Here, we employ a combination of statistical sequence analysis, molecular dynamics simulations, and in vitro mutational analysis to define hallmarks of DCLK family evolutionary divergence, including analysis of splice variants within the DCLK1 sub-family, which arise through alternative codon usage and serve to 'supercharge' the inhibitory potential of the DCLK1 C-tail. We identify co-conserved motifs that readily distinguish DCLKs from all other calcium calmodulin kinases (CAMKs), and a 'Swiss Army' assembly of distinct motifs that tether the C-terminal tail to conserved ATP and substrate-binding regions of the catalytic domain to generate a scaffold for autoregulation through C-tail dynamics. Consistently, deletions and mutations that alter C-terminal tail length or interfere with co-conserved interactions within the catalytic domain alter intrinsic protein stability, nucleotide/inhibitor binding, and catalytic activity, suggesting isoform-specific regulation of activity through alternative splicing. Our studies provide a detailed framework for investigating kinome-wide regulation of catalytic output through cis-regulatory events mediated by intrinsically disordered segments, opening new avenues for the design of mechanistically divergent DCLK1 modulators, stabilizers, or degraders.