Variants of intrinsic disorder in the human proteome

In this paper we propose a straightforward operational definition of variants of disordered proteins, taking the human proteome as a case study. The focus is on a distinction between mostly unstructured proteins and proteins which contain long unstructured regions accommodated in an overall folded structure. In particular we distinguish: i) Not disordered proteins (NDPs), that either have all their residues ordered or do not have disordered segments longer than 30 residues nor more than 30% of disordered residues; ii) Proteins with intrinsically disordered regions (IDRPs), that have at least one disordered domain longer than 30 residues, but disordered in less than 30% of their residues; iii) Proteins that are intrinsically disordered (IDPs), that have both at least one disordered segment longer than 30 residues and that are disordered on more than 30% of their residues; iv) Proteins with fragmented disorder (FRAG_IDPs), that do not have a disordered fragment longer than 30 residues but that, nevertheless, have at least 30% or more of their residues predicted as disordered. The potential use of these variants is checked over several groups of disease-related proteins. Our main conclusions point out that IDRPs are more similar to NDPs than to IDPs. IDRPs and NDPs have a similar functional repertoire and probably share a lock-and-key mechanism of interaction with substrates. IDRPs and IDPs are differently present among human disease-related proteins. IDRPs probably do not play a specific role in the development of complex diseases, since their frequency is similar in disease-related proteins and in the entire human proteome. IDPs can play a role in the emergence of cancer, neurodegenerative, thyroid and liver disease.