What these predictions describe
These predictions capture the emergent properties of a protein — the biophysical propensities encoded in its sequence — rather than the behaviour of one final folded state. That distinction matters here: proteins are dynamic even when folded, and some do not fold at all, as with intrinsically disordered proteins.
To run these same predictors on a sequence of your own, use the Bio2Byte online predictors, or read about the underlying tools on our website.
The predictors
In the order the atlas lists them. Every reference opens at the publisher.
DynaMine
Backbone and side-chain dynamics
-
From
protein sequence to dynamics and disorder with DynaMine
Nature Communications 4, 3741 (2013). -
The
DynaMine webserver: predicting protein dynamics from sequence
Nucleic Acids Research 42, W264–W270 (2014).
DisoMine
Disorder
-
Prediction
of disordered regions in proteins with recurrent neural networks and
protein dynamics
Journal of Molecular Biology 434, 167579 (2022).
EFoldMine
Early folding
-
Exploring
the sequence-based prediction of folding initiation sites in
proteins
Scientific Reports 7, 8826 (2017).
AgMata
Beta-sheet aggregation
-
Accurate
prediction of protein beta-aggregation with generalized statistical
potentials
Bioinformatics 36, 2076–2081 (2020).
PSPer
Phase separation
-
Computational
identification of prion-like RNA-binding proteins that form liquid
phase-separated condensates
Bioinformatics 35, 4617–4623 (2019).
Biophysical interpretation of disorder
Random Forest classifier
-
Challenges
in describing the conformation and dynamics of proteins with ambiguous
behavior
Frontiers in Molecular Biosciences 9, 959956 (2022).
This one is an additional classifier rather than part of b2bTools. It reads the predictions above — backbone and side-chain dynamics, the conformational propensities, early folding and disorder — and labels each residue as order, transition or disorder, so it is available for every entry that carries those predictions.
Sequences and scope
Sequences come from UniProt's reference proteomes. This atlas covers the reviewed entries of each one — the manually curated Swiss-Prot section of UniProtKB — so results are comparable between proteins and between species. Each proteome's own page states how much of it that is.
Predictions are computed per residue on the canonical sequence, with no structural input, which is why they are available for every entry rather than only those with a solved structure.