henhaobio
Sign in

bioRxivBioinformaticsPreprint

Porter 5: fast, state-of-the-art ab initio prediction of protein secondary structure in 3 and 8 classes

Mirko TorrisiiD, Manaz KaleeliD, Gianluca PollastriiD

PDFDOI
Structured summary available · Jump to summary ↓

Abstract

MotivationAlthough Secondary Structure Predictors have been developed for more than 60 years, current ab initio methods have still some way to go to reach their theoretical limits. Moreover, the continuous effort towards harnessing ever increasing data sets and more sophisticated, deeper Machine Learning techniques, has not come to an end.\n\nResultsHere we present Porter 5, the last release of one of the best performing ab initio secondary structure predictor. Version 5 achieves 84% accuracy (84% SOV) when tested on 3 classes, and 73% accuracy (82% SOV) on 8 classes, on a large independent set, significantly outperforming all the most recent ab initio predictors we have tested.\n\nAvailabilityThe web and standalone versions of Porter5 are available at http://distilldeep.ucd.ie/porter.\n\nContactname@bio.com\n\nSupplementary informationSupplementary data are available at Bioinformatics online.

Abstract summary

Claude Opus 5.5 · xhigh effort · 2 Nov 2026 · from the abstract

Porter 5 predicts ab initio protein secondary structure at 84% 3-class accuracy and 73% 8-class accuracy on an independent test set, available as web and standalone tools.

Despite more than 60 years of development, ab initio protein secondary structure predictors have not yet reached their theoretical accuracy limits and continue to benefit from larger datasets and deeper machine learning techniques. Porter 5 updates the Porter ab initio predictor to improve 3-class and 8-class secondary structure prediction accuracy over recent competing methods.

Key results

  1. ShownPorter 5 predicts 3-class and 8-class protein secondary structure with higher accuracy than other recent ab initio predictors tested on an independent dataset.

    Achieved 84% accuracy and 84% segment overlap (SOV) on 3 classes, and 73% accuracy and 82% SOV on 8 classes on an independent test set; specific competing predictor scores and test set size are not reported in the abstract. · Abstract

Methods, data and reproducibility

The authors developed Porter 5, a machine-learning-based ab initio protein secondary structure predictor for 3-class and 8-class assignments, and benchmarked its accuracy and segment overlap (SOV) on an independent test set against recent ab initio predictors. Web and standalone versions were released.

  • Independent test set of proteins for evaluating 3-class and 8-class secondary structure prediction
Methods & tools
Porter 5
Reproducibility gaps
Training and test dataset accessions, model architecture details, and compute requirements are not provided in the abstract.
Limitations (1)
  • Abstract only; neural network architecture, training set composition, independent test set size, and numerical scores of competing predictors are not stated. · noted in this summary

Read this if you need a web or standalone ab initio predictor for 3-class or 8-class protein secondary structure from sequence.

Citations
46

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Related papers

Similarity Map
SelectedSame fieldCross-field
THISPeracha '23Uddin '20Hou '19Bliss '19Shapovalov '20Klausen '19
67% semantic matchBioinformatics
Omar Peracha·BioTechniques·Nov 24, 2023· 7 citations