KinoPlex, a computational framework that integrates predicted protein structures and kinase recognition motifs to assign phosphorylation potential and kinase specificity to all serine/threonine/tyrosine residues, is presented.
Protein kinases regulate cell signaling through phosphorylation of serine, threonine, or tyrosine residues on substrate proteins. Their catalytic activity is governed by a set of conserved structural elements, the activation segment (bounded by the DFG and APE motifs), the DFG motif, the activation loop (A-loop), and the αC-helix, whose conformational states determine whether a kinase is active or inactive. Despite substantial efforts to classify kinase conformations, most existing schemes are geometric in nature; few integrate a quantitative energetic dimension, and the finer secondary-structure features of the activation segment remain underexploited. We compiled a dataset of 4,670 human and murine protein kinase structures from the RCSB PDB (1,699 tyrosine kinases and 2,971 serine/threonine kinases). Activation segment configurations (IN, OUT, and SWAPPED) were assigned using DSSP-guided structural inspection and geometric criteria. Association diagrams were built for 104 tyrosine kinases and 110 serine/threonine kinases. For 68 catalytic domains with fully resolved activation segments, K-means clustering was applied using four criteria: activation segment interaction energy (INTAA server, AMBER parm03 force field), and the conformational states of the DFG motif, αC-helix, and A-loop. Spatial heat maps and per-residue Cα displacement analysis (VMD) were used to characterize energy distribution and conformational transitions. The activation segment was classified into OUT (active, 55%), IN (inactive, 38%), and a minor SWAPPED conformation (6% in the broad survey; n = 2 in the fully resolved clustering subset), the latter retained as an observation rather than a validated class. K-means clustering defined seven descriptive energy/conformation clusters. Heat maps revealed that the activation segment interacts primarily with the catalytic loop, the β1 strand, and the αEF/αF loop, with varying intensities across clusters. Per-residue Cα displacement analysis showed that the A-loop undergoes the largest conformational change between inactive and active states (up to 23 Å), with smaller kinase-specific differences involving the G-loop, αC-helix, and adjacent β-strands. This study provides an integrated energetic and structural classification of the activation segment across the human and murine kinome, complementing existing conformational catalogues and offering a quantitative comparative basis for understanding kinase activation mechanisms relevant to drug design.
A. Ahiri, A. Aboulmouhajir· The European Chemistry and B...· 0 citations
A pipeline reformulating kinase-substrate modeling as a Bayesian inference problem is presented and it is revealed that the interaction types and distances to the catalytic pocket significantly influence pathogenicity scores.
Jinyuan Hu, Shimian Li, Yue Xue et al.· Journal of Chemical Informat...· 0 citations
Mass-spectrometry-based phosphoproteomics now profiles phosphorylation at proteome scale, yet converting site-level measurements into coherent, kinase-centered biology — and into actionable drug discovery decisions — remains a persistent barrier to target nomination, mechanism-of-action confirmation, and resistance management. The Kinase Library addresses this gap with the first-in-class, unbiased, experimentally characterized motif atlas of the human kinome, coupled to enrichment frameworks that translate phosphoproteomics data into quantitative, rank-ordered maps of kinase activity. Rather than relying on heterogeneous annotations or heuristic rules, the Kinase Library grounds inference in experimentally derived kinase-substrate relationships, providing a principled basis for target deconvolution, mechanism-of-action analysis, and comparative pharmacology. The Kinase Library has broad utility across drug discovery and development. It enables on- and off-target mechanism-of-action profiling for small molecules and combinations; delineates adaptive signaling and resistance trajectories that drive clinical relapse; supports time-course and dose-response studies to resolve pathway dynamics and therapeutic windows; nominates rational combination partners by pairing on-target deconvolution with kinome-wide compensatory readouts; and stratifies models and patients in low-N-high-D (few samples with high dimensionality of data) settings where conventional statistics underperform. In preclinical and clinical contexts alike — cell lines, organoids, xenografts, and patient specimens — the Kinase Library delivers harmonized, interpretable kinase signatures that integrate readily with multiomic readouts to generate, prioritize, and de-risk actionable drug-development hypotheses. The novelty of the Kinase Library is twofold. First, scope and provenance: an experimental, unbiased atlas spanning the entire kinome, with comprehensive inclusion of the dark kinome — opening previously inaccessible target space for medicinal chemistry. Second, operationalization: a unified enrichment paradigm that yields robust, rank-ordered kinase programs suitable for go/no-go decision-making — whether the objective is target nomination, combination design, biomarker discovery, patient stratification, or comparative benchmarking across cohorts, modalities, and studies. Looking forward, the Kinase Library is positioned to empower emerging frontiers across the drug-discovery continuum: single-cell and spatial phosphoproteomics; longitudinal "N-of-1" pharmacodynamic monitoring to guide therapy selection; cross-species translation for preclinical model qualification; and cloud-native workflows that interoperate with community pipelines, public datasets, and pharma-internal infrastructure. By elevating kinases from disparate lists of regulated sites to coherent, testable signaling hypotheses, the Kinase Library reframes what phosphoproteomics can deliver — shifting the field from descriptive measurement toward predictive, mechanism-guided drug discovery and intervention.
Tomer M. Yaron-Barir, Jared L. Johnson, Lewis C. Cantley. The Kinase Library: A global atlas of the human protein kinome and its applications in drug discovery [abstract]. In: Proceedings of AACR Drug Discovery and Development (AACR D3) Conference; 2026 Jul 21-24; Boston, MA. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(14_Suppl):Abstract nr B052.
Tomer M. Yaron-Barir, Jared L. Johnson, Lewis C. Cantley· Clinical Cancer Research· 0 citations
Protein kinases are critical drug targets, requiring therapeutics that can modulate their active and inactive conformational states. While cofolding models can generate global folds directly from kinase sequences and ligand SMILES strings, these models have not yet been tested on their ability to recover ligand-induced-fit conformational states of the kinase proteins. Here, we introduce KinConfBench, a curated benchmark of 2225 high-quality human kinase chains to evaluate the ability of four state-of-the-art cofolding models—Boltz-2, Chai-1, Protenix, and RoseTTAFold-All-Atom—to recover both canonical and rare conformational states. We show that geometric success metrics of a ligand pose in the active site do not correlate strongly with the correct kinase conformational state, motivating a new set of dynamical benchmarks for assessing cofolding models. While all four cofolding models achieve ~60–80% prediction accuracy for kinase conformational classification, they exhibit severe mode collapse when performing multiple inferences, show negligible structural diversity in sampling induced-fit motions, and display a prevalent “apo-drift” in which most cofolding models predominantly predict the kinase to be in its ligand-free state. Our results highlight that capturing ligand-induced protein conformational diversity, not just geometric fit, is critical for next-generation structure-based drug discovery.
Kunyang Sun, T. Head-Gordon· npj Drug Discovery· 0 citations
The regulation of protein stability is essential for cellular homeostasis and is determined by a combination of intrinsic sequence motifs and extrinsic recognition enzymes. Despite growing knowledge of the protein degradation machinery, the ability to predict a protein’s stability from its amino acid sequence remains challenging. Here we develop a machine learning model to predict protein stability from N-terminal amino acid sequences. Using our model and experimental validation, we identify known and novel sequence motifs governing protein stability. We additionally use this model to predict the stability of alternative translational isoforms with distinct N-termini produced from the same mRNA. Despite differing by a limited number of amino acids, we identify N-terminal isoforms with drastically different stabilities relative to their annotated counterparts, highlighting the potential of N-terminal extensions and truncations to regulate protein function. Together, this model provides a valuable tool for evaluating additional protein datasets and protein design strategies.
Océane Marescal, Iain M. Cheeseman· bioRxiv· 0 citations
A structure-based method using SE(3)-transformers to learn residue compatibility with the local structural environment from experimentally resolved kinase 3D structures that captures biologically meaningful relationships between residue identity and 3D structural context is presented.
Shakiba Fadaei, F. Krebs, V. Zoete· Bioinformatics· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
