Predictomes, a classifier-curated database of AlphaFold-modeled protein-protein interactions

Molecular Cell, Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 1, 2025

Protein-protein interactions (PPIs) are ubiquitous in biology, yet a comprehensive structural characterization of the PPIs underlying cellular processes is lacking. AlphaFold-Multimer (AF-M) has potential to fill this knowledge gap, but standard AF-M confidence metrics do not reliably separate relevant from an abundance false positive predictions. To address limitation, we used machine learning on curated datasets train structure prediction and omics-informed classifier (SPOC) that effectively separates true predictions PPIs, including proteome-wide screens. We applied SPOC all-by-all matrix nearly 300 human genome maintenance proteins, generating ∼40,000 can be viewed at predictomes.org, where users also score their own with SPOC. High-confidence discovered using our approach enable hypothesis generation maintenance. Our results provide framework for interpreting large-scale screens help lay foundation interactome.

Language: Английский

---

Rēs ipSAE loquunt: What′s wrong with AlphaFold′s ipTM score and how to fix it

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 14, 2025

AlphaFold's ipTM metric is used to predict the accuracy of structural predictions protein-protein interactions (PPIs) and probability that two proteins interact. Many AF2/AF3 users have experienced phenomenon if they trim full-length sequence constructs (e.g. from UniProt) interacting domains (or domain+peptide), their scores go up, even though structure prediction interaction unchanged. The reason this happens due mathematical formulation in AF2/AF3, which whole chains. If both chains a PPI complex contain large amounts disorder or accessory do not form primary domain-domain domain/peptide interaction, score can be lowered significantly. then does accurately represent nor whether actually We solved problem by: 1) including only residue pairs good predicted aligned error ( PAE ) scores; 2) by adjusting d 0 parameter (a function length query sequences) TM equation include number residues with interchain s residue; 3) using value itself distributions over calculate pairwise residue-residue pTM values into calculation. first are crucial calculating high for domain-peptide presence many hundreds disordered regions and/or domains. third allows us require common output json files AF2 AF3 (including server output) without having change AlphaFold code affecting accuracy. show benchmark new score, called ipSAE (interaction Score Aligned Errors), able separate true false complexes more efficiently than AlphaFold2's score. resulting program freely available at https://github.com/dunbracklab/IPSAE .

Language: Английский

---

Investigating plasticity within the interleukin-6 family with AlphaFold-Multimer

Computational and Structural Biotechnology Journal, Journal Year: 2025, Volume and Issue: 27, P. 946 - 959

Published: Jan. 1, 2025

Cytokines are important soluble mediators that involved in physiological and pathophysiological processes. Among them, members of the interleukin-6 (IL-6) family cytokines have gained remarkable attention, because especially name-giving cytokine IL-6 has been shown to be an excellent target treat inflammatory autoimmune diseases. The consists nine members, which activate their cells via combinations non-signaling α- and/or signal-transducing β-receptors. While some receptor exclusively used by a single cytokine, other multiple cytokines. Research recent years unraveled another level complexity: several cannot only signal canonical receptors, but can bind additional β-receptors, albeit with less affinity. examples such plasticity reported, systematic analysis this phenomenon is lacking. development artificial intelligence programs like AlphaFold allows computational protein complexes manner. Here, we develop pipeline for cytokine:cytokine interaction show AlphaFold-Multimer correctly predicts ligands family. However, does not provide sufficient insight conclusively predict alternative, low-affinity receptors within

Language: Английский

---

Engineered Proteins and Chemical Tools to Probe the Cell Surface Proteome

Chemical Reviews, Journal Year: 2025, Volume and Issue: unknown

Published: April 3, 2025

The cell surface proteome, or surfaceome, is the hub for cells to interact and communicate with outside world. Many disease-associated changes are hard-wired within yet approved drugs target less than 50 proteins. In past decade, proteomics community has made significant strides in developing new technologies tailored studying surfaceome all its complexity. this review, we first dive into unique characteristics functions of emphasizing necessity specialized labeling, enrichment, proteomic approaches. An overview surfaceomics methods provided, detailing techniques measure protein expression how leads novel discovery. Next, highlight advances proximity labeling (PLP), showcasing various enzymatic photoaffinity can map protein-protein interactions membrane complexes on surface. We then review role extracellular post-translational modifications, focusing glycosylation, proteolytic remodeling, secretome. Finally, discuss identifying tumor-specific peptide MHC they have shaped therapeutic development. This emerging field neo-protein epitopes constantly evolving, where targets identified at proteome level encompass defined PTMs, complexes, dysregulated cellular tissue locations. Given functional importance biology therapy, view as a critical piece quest neo-epitope

Language: Английский

---

Crystal structure of Isthmin-1 and reassessment of its functional role in pre-adipocyte signaling

Nature Communications, Journal Year: 2025, Volume and Issue: 16(1)

Published: April 15, 2025

Isthmin-1 (ISM1) is a recently described adipokine with insulin-like properties that can control hyperglycemia and liver steatosis. Additionally, ISM1 proposed to play critical roles in patterning, angiogenesis, vascular permeability, apoptosis. A key feature of its AMOP (adhesion-associated domain MUC4 (Mucin-4) other proteins) which essential for many functions. However, the molecular details domains remain elusive as there are no descriptions their structure. Here we determined crystal structure including thrombospondin type I repeat (TSR) domain. Interestingly, ISM1's exhibits distinct fold similarities bacterial streptavidin. When comparing our predicted structures domains, observed while core streptavidin-like barrel conserved, surface helices loops vary greatly. Thus, allows structural plasticity may underpin diverse Furthermore, contrary prior studies, show highly purified does not stimulate AKT phosphorylation on 3T3-F442A pre-adipocytes. Rather, find co-purifying growth factors responsible this activity. Together, data reveal clarify functional studies enigmatic protein.

Language: Английский

---

Emerging frontiers in protein structure prediction following the AlphaFold revolution

Journal of The Royal Society Interface, Journal Year: 2025, Volume and Issue: 22(225)

Published: April 1, 2025

Models of protein structures enable molecular understanding biological processes. Current structure prediction tools lie at the interface biology, chemistry and computer science. Millions models have been generated in a very short space time through revolution driven by deep learning, led AlphaFold. This has provided wealth new structural information. Interpreting these predictions is critical to determining where when this information useful. But proteins are not static nor do they act alone, interacting with other biomolecules complete their function level. review focuses on application state-of-the-art advanced applications. We also suggest set guidelines for reporting AlphaFold predictions.

Language: Английский

---

PPIscreenML: Structure-based screening for protein-protein interactions using AlphaFold

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: March 17, 2024

Abstract Protein-protein interactions underlie nearly all cellular processes. With the advent of protein structure prediction methods such as AlphaFold2 (AF2), models specific pairs can be built extremely accurately in most cases. However, determining relevance a given pair remains an open question. It is presently unclear how to use best structure-based tools infer whether candidate proteins indeed interact with one another: ideally, might even information screen amongst pairings build up interaction networks. Whereas for evaluating quality modeled complexes have been co-opted which (e.g., pDockQ and iPTM), there no rigorously benchmarked this task. Here we introduce PPIscreenML, classification model trained distinguish AF2 interacting from compelling decoy pairings. We find that PPIscreenML out-performs iPTM task, further exhibits impressive performance when identifying ligand/receptor engage another across structurally conserved tumor necrosis factor superfamily (TNFSF). Analysis benchmark results using not seen development strongly suggest generalizes beyond training data, making it broadly applicable new based on structural AF2.

Language: Английский

---

AlphaFold opens the doors to deorphanizing secreted proteins

Cell Systems, Journal Year: 2024, Volume and Issue: 15(11), P. 1000 - 1001

Published: Nov. 1, 2024

---