Assessment of hard target modeling in CASP12 reveals an emerging role of alignment‐based contact prediction methods

Abriata, Luciano A.; Tamò, Giorgio E.; Monastyrskyy, Bohdan; Kryshtafovych, Andriy; Peraro, Matteo Dal

doi:10.1002/prot.25423

Cited by 87 publications

(105 citation statements)

References 52 publications

Supporting

Mentioning

101

Contrasting

Order By: Relevance

“…In order to check whether the results of the methods were significantly different, we calculated a P value for each feature by using a pairwise Student's t test on the results of the 2 methods. Results on the independent test sets CASP12, TS115, and CB513 are presented in Table , and in an extended version in Supporting Information Table S1. They match to a very high degree with the results obtained by using a 4‐fold cross validation on the training set (Supporting Information Table S2).…”

Section: Resultsmentioning

confidence: 99%

NetSurfP‐2.0: Improved prediction of protein structural features by integrated deep learning

et al. 2019

View full text Add to dashboard Cite

The ability to predict local structural features of a protein from the primary sequence is of paramount importance for unraveling its function in absence of experimental structural information. Two main factors affect the utility of potential prediction tools: their accuracy must enable extraction of reliable structural information on the proteins of interest, and their runtime must be low to keep pace with sequencing data being generated at a constantly increasing speed. Here, we present NetSurfP‐2.0, a novel tool that can predict the most important local structural features with unprecedented accuracy and runtime. NetSurfP‐2.0 is sequence‐based and uses an architecture composed of convolutional and long short‐term memory neural networks trained on solved protein structures. Using a single integrated model, NetSurfP‐2.0 predicts solvent accessibility, secondary structure, structural disorder, and backbone dihedral angles for each residue of the input sequences. We assessed the accuracy of NetSurfP‐2.0 on several independent test datasets and found it to consistently produce state‐of‐the‐art predictions for each of its output features. We observe a correlation of 80% between predictions and experimental data for solvent accessibility, and a precision of 85% on secondary structure 3‐class predictions. In addition to improved accuracy, the processing time has been optimized to allow predicting more than 1000 proteins in less than 2 hours, and complete proteomes in less than 1 day.

show abstract

Section: Resultsmentioning

confidence: 99%

NetSurfP‐2.0: Improved prediction of protein structural features by integrated deep learning

et al. 2019

View full text Add to dashboard Cite

show abstract

“…Performance vs CASP round was plotted from the CASP12 plot and CASP13 data. All the data presented in the web app, HHpred (computed as HHblits score × coverage) and LGA score presented here, Neff, Grishin plots, were provided by the Prediction Center.…”

Section: Methodsmentioning

confidence: 99%

A further leap of improvement in tertiary structure prediction in CASP13 prompts new routes for future assessments

2019

Self Cite

View full text Add to dashboard Cite

We present our assessment of tertiary structure predictions for hard targets in Critical Assessment of Structure Prediction round 13 (CASP13). The analysis includes (a) assignment and discussion of best models through scores‐aided visual inspection of models for each evaluation unit (EU); (b) ranking of predictors resulting from this evaluation and from global scores; and (c) evaluation of progress, state of the art, and current limitations of protein structure prediction. We witness a sizable improvement in tertiary structure prediction building on the progress observed from CASP11 to CASP12, with (a) top models reaching backbone RMSD <3 å for several EUs of size <150 residues, contributed by many groups; (b) at least one model that roughly captures global topology for all EUs, probably unprecedented in this track of CASP; and (c) even quite good models for full, unsplit targets. Better structure predictions are brought about mainly by improved residue‐residue contact predictions, and since this CASP also by distance predictions, achieved through state‐of‐the‐art machine learning methods which also progressed to work with slightly shallower alignments compared to CASP12. As we reach a new realm of tertiary structure prediction quality, new directions are proposed and explored for future CASPs: (a) dropping splitting into EUs, (b) rethinking difficulty metrics probably in terms of contact and distance predictions, (c) assessing also side chains for models of high backbone accuracy, and (d) assessing residue‐wise and possibly residue‐residue quality estimates.

show abstract

“…By comparing residue-residue coevolution strengths computed from an alignment of Cyc2-like proteins with the Gremlin server, 7,8 it appears that the predicted contacts match best with the contacts present in model #3 (Fig. 2B and different cutoffs of the coevolution data in Fig.…”

Section: Resultsmentioning

confidence: 99%

Structural model of Cyc2, the primary electron acceptor of Acidithiobacillus ferrooxidans’ respiratory chain, as a modular cytochrome - β-barrel fusion protein, and mechanistic proposals based on this model

Abriata

2018

Preprint

Self Cite

View full text Add to dashboard Cite

Abstract. Acidithiobacillus ferrooxidans oxidizes Fe(II) to Fe(III) to feed electrons into its respiratory chain. The primary electron acceptor of this complex system is Cyc2, an outer membrane protein of unknown structure. This work proposes a feasible model of Cyc2's global structure, based on homology modeling, residue-residue coevolution data, bioinformatics predictions and limited knowledge about Cyc2's function. The proposal is that the sequence segment spanning residues ~30 to ~90 folds as a cytochrome-like domain that contains a heme group which would presumably bind and oxidize external Fe(II), whereas the remaining segment from residue ~90 until the end adopts a β-barrel fold similar to that of most outer membrane proteins. Such model differs strongly from a published model, but is backed up by more data and is more compatible with the known topology of outer membrane proteins and with Cyc2's function of internalizing reducing equivalents. The small size of the cytochrome-like domain would allow it to reside inside, and/or slide through, the β-barrel domain, thus communicating in a controlled fashion the extracellular medium with the periplasm to import electrons through the outer membrane. All the models discussed are provided as PyMOL session files in the Supporting Information and can be visualized online at

show abstract

Assessment of hard target modeling in CASP12 reveals an emerging role of alignment‐based contact prediction methods

Cited by 87 publications

References 52 publications

NetSurfP‐2.0: Improved prediction of protein structural features by integrated deep learning

NetSurfP‐2.0: Improved prediction of protein structural features by integrated deep learning

A further leap of improvement in tertiary structure prediction in CASP13 prompts new routes for future assessments

Structural model of Cyc2, the primary electron acceptor of Acidithiobacillus ferrooxidans’ respiratory chain, as a modular cytochrome - β-barrel fusion protein, and mechanistic proposals based on this model

Contact Info

Product

Resources

About