DFLpred: High-throughput prediction of disordered flexible linker regions in protein sequences

Abstract

Motivation: Disordered flexible linkers (DFLs) are disordered regions that serve as flexible linkers/spacers in multi-domain proteins or between structured constituents in domains. They are different from flexible linkers/residues because they are disordered and longer. Availability of experimentally annotated DFLs provides an opportunity to build high-throughput computational predictors of these regions from protein sequences. To date, there are no computational methods that directly predict DFLs and they can be found only indirectly by filtering predicted flexible residues with predictions of disorder.Results: We conceptualized, developed and empirically assessed a first-of-its-kind sequence-based predictor of DFLs, DFLpred. This method outputs propensity to form DFLs for each residue in the input sequence. DFLpred uses a small set of empirically selected features that quantify propensities to form certain secondary structures, disordered regions and structured regions, which are processed by a fast linear model. Our high-throughput predictor can be used on the whole-proteome scale; it needs <1 h to predict entire proteome on a single CPU. When assessed on an independent test dataset with low sequence-identity proteins, it secures area under the receiver operating characteristic curve equal 0.715 and outperforms existing alternatives that include methods for the prediction of flexible linkers, flexible residues, intrinsically disordered residues and various combinations of these methods. Prediction on the complete human proteome reveals that about 10% of proteins have a large content of over 30% DFL residues. We also estimate that about 6000 DFL regions are long with ≥30 consecutive residues.Availability and implementation: http://biomine.ece.ualberta.ca/DFLpred/.Contact: lkurgan@vcu.eduSupplementary information: Supplementary data are available at Bioinformatics online.

How disordered is my protein and what is its disorder for? A guide through the “dark side” of the protein universe

“…95,96 One of the most recent methods, DFLpred, predicts disordered regions that serve as either intra-domain or inter-domain linkers. 97 Finally, it has been reported that sites of the enzyme-catalyzed posttranslational modifications, such as phosphorylation, 98 acetylation, methylation, and ubiquitination 99 are commonly located within the IDRs. Several computational tools utilizing this information have been developed, such as DisPhos (Disorder-enhanced Phosphorylation predictor), which can efficiently find IDR-located phosphorylation sites with 76% accuracy for serine, 81% for threonine, and 83% for tyrosine.…”

High‐throughput prediction of disordered moonlighting regions in protein sequences

“…The length of the second type of the windows varies and is determined by the length of the putative disordered regions generated with IUPred_short and IUPred_long. The use of the fixed size sliding windows is motivated by the design of related methods, such as MoRFpred, fMoRFpred, DisoRDPbind, and DFLpred . Using the individual and combined biophysical and structural properties that are quantified for individual residues and based on the two types of windows, we compute 1588 features for each residue in the input protein chain.…”

“…The use of the fixed size sliding windows is motivated by the design of related methods, such as MoRFpred, 24 fMoRFpred, 25 DisoRDPbind, 28 and DFLpred. 27 Using the individual and combined biophysical and structural properties that are quantified for individual residues and based on the two types of windows, we compute 1588…”

“…Moreover, progress has been made in recent years to predict functions of IDRs . Example predictors include Anchor, MoRFpred, fMoRFpred, and MoRFCHiBi that predict disordered protein–protein binding regions, DFLpred that outputs putative disordered linkers, and DisoRDPbind that predicts disordered protein–protein, RNA‐ and DNA‐binding regions. High degree of plasticity of IDRs allows them to bind more than one ligand and carry out multiple functions .…”

Partner‐specific prediction of RNA‐binding residues in proteins: A critical assessment

“…81 Recently, IDR have been successfully used for predicting RNA and DNA-binding proteins on a large scale. 82 This finding, together with the availability of increasing number of computational tools for predicting IDRs from protein sequence, [83][84][85] suggests a promising direction for developing improved protein-RNA binding site prediction models that incorporate information derived from IDRs. Another promising direction is to leverage high throughput data from protein-RNA binding assays and affinity distributions, to generate the RNA binding models for RBPs.…”