Biology subjects

Ferrie, J. J.

Publications and source records attributed to Ferrie, J. J..

5 recordsLinked to original sources

bioRxiv2023

Prot-SCL: State Of The Art Prediction Of Protein Subcellular Localization From Primary Sequence Using Contrastive Learning

Protein subcellular localization is a critically important parameter to consider when designing expression constructs and production strategies for industry scale protein production. In this study, we present Prot-SCL an innovative self-supervised machine learning approach to predict protein subcellular localization exclusively from primary sequence. The models herein were learned from a dataset of subcellular localizations derived by exhaustively analyzing the Uniprot database. The set of localization data was rigorously curated for machine learning by employing group sampling following clustering of the protein sequences. The novel component of this approach lies in the development of a triplet neural network architecture capable of generating meaningful embeddings for classification of protein subcellular localization. We observed a robust predictive power for our classical gradient boosted machine learning models trained on these triplet embeddings in both cross validation and in generalization to the testing set. Importantly, we have made this extensive dataset of protein subcellular localizations publicly accessible, facilitating future, need-based, localization studies. Finally, we provide the relevant codebase to encourage a wider adoption and expansion of this methodology. GRAPHICAL ABSTRACT O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=123 SRC="FIGDIR/small/555932v1_ufig1.gif" ALT="Figure 1"> View larger version (52K): org.highwire.dtl.DTLVardef@1d9b205org.highwire.dtl.DTLVardef@1368c23org.highwire.dtl.DTLVardef@2a88edorg.highwire.dtl.DTLVardef@8380b7_HPS_FORMAT_FIGEXP M_FIG C_FIG

bioinformatics↗

bioRxiv2023

p300 Is an Obligate Integrator of Combinatorial Transcription Factor Inputs

Transcription coactivators are proteins or protein complexes that mediate transcription factor (TF) function. However, they lack DNA binding capacity, prompting the question of how they engage target loci. Three non-exclusive hypotheses have been posited: coactivators are recruited by complexing with TFs, by binding histones through epigenetic reader domains, or by partitioning into phase-separated compartments through their extensive intrinsically disordered regions (IDRs). Using p300 as a prototypical coactivator, we systematically mutated its annotated domains and show by single-molecule tracking in live cells that coactivator- chromatin binding depends entirely on combinatorial binding of multiple TF-interaction domains. Furthermore, we demonstrate that acetyltransferase activity negatively impacts p300-chromatin association and that the N-terminal TF-interaction domains regulate that activity. Single TF-interaction domains are insufficient for both chromatin binding and regulation of catalytic activity, implying a principle that could broadly inform eukaryotic gene regulation: a TF must act in coordination with other TFs to recruit coactivator activity.

biophysics↗

bioRxiv2021

Proximity-assisted photoactivation (PAPA): Detecting molecular interactions in live-cell single-molecule imaging

Single-molecule imaging provides a powerful way to study biochemical processes in live cells, yet it remains challenging to track single molecules while simultaneously detecting their interactions. Here we describe a novel property of rhodamine dyes, proximity-assisted photoactivation (PAPA), in which one fluorophore (the "sender") can reactivate a second fluorophore (the "receiver") from a dark state. PAPA requires proximity between the two fluorophores, yet it operates at a longer average intermolecular distance than Forster resonance energy transfer (FRET). We show that PAPA can be used in live cells both to detect protein-protein interactions and to highlight a sub-population of labeled protein complexes in which two different labels are in proximity. In proof-of-concept experiments, PAPA detected the expected correlation between androgen receptor self-association and chromatin binding at the single-cell level. These results establish a new way in which a photophysical property of fluorophores can be harnessed to study molecular interactions in single-molecule imaging of live cells.

biophysics↗

bioRxiv2020

Rosetta Custom Score Functions Accurately Predict ΔΔG of Mutations at Protein-Protein Interfaces Using Machine Learning

Protein-protein interfaces play essential roles in a variety of biological processes and many therapeutic molecules are targeted at these interfaces. However, accurate predictions of the effects of interfacial mutations to identify "hotspots" have remained elusive despite the myriad of modeling and machine learning methods tested. Here, for the first time, we demonstrate that nonlinear reweighting of energy terms from Rosetta, through the use of machine learning, exhibits improved predictability of {Delta}{Delta}G values associated with interfacial mutations.

biophysics↗

bioRxiv2020

A Unified De Novo Approach for Predicting the Structures of Ordered and Disordered Proteins

As recognition of the abundance and relevance of intrinsically disordered proteins (IDPs) continues to grow, demand increases for methods that can rapidly predict the conformational ensembles populated by these proteins. To date, IDP simulations have largely been dominated by molecular dynamics (MD) simulations, which require significant compute times and/or complex hardware. Recent developments in MD have afforded methods capable of simulating both ordered and disordered proteins, yet to date accurate fold prediction from sequence has been dominated by Monte-Carlo (MC) based methods such as Rosetta. To overcome the limitations of current approaches in IDP simulation using Rosetta while maintaining its utility for modeling folded domains, we developed PyRosetta-based algorithms that allow for the accurate de novo prediction of proteins across all degrees of foldedness along with structural ensembles of disordered proteins. Our simulations have an accuracy comparable to state-of-the-art MD with vastly reduced computational demands.

biophysics↗