IDDomainSpotter is a sequence-based approach to assess and visualize domain organisation in long intrinsically disordered proteins based on compositional biases. IDDomainSpotter calculates the fractions of residue types in either single protein sequences or sequence alignments, in sliding windows with a default size of 15. For each residue, k, the fraction of residues between k-7 to k+7 matching specific criteria is returned. All possible criteria and combinations can be applied, and the user can add as many features as desired, as well as choose a custom window size relevant to the user’s study. Each feature consists of a specific combination of amino acids yielding a positive contribution and a combination of amino acids yielding a negative contribution. For each feature any combination of proteinogenic amino acids (or X for unusual or unnatural amino acids) can be used, making the approach highly flexible and customizable depending on the protein examined and the properties of choice.
You find IDDomainSpotter here: http://www2.bio.ku.dk/IDR/.
IDDomainSpotter does not predict disorder, so the user is referred to other tools for this. When analysing composite features, we suggest careful consideration of individual amino acid contributions, e.g. by also observing the profiles of each individual amino acid. It is crucial to pay special attention when using IDDomainSpotter with sequence alignments, as poor or unbalanced alignments may yield artefacts or mask genuine signals.
Millard, P. S., Bugge, K., Marabini, R., Boomsma, W., Burow, M., Kragelund, B. B. (2019) IDDomainSpotter: Compositional bias reveals domains in long disordered protein regions-Insights from transcription factors. Protein Sci. DOI: 10.1002/pro.3754, PMID: 31642121