An expanded evaluation of protein function prediction methods shows an improvement in accuracy.
Yuxiang JiangTal Ronnen OronWyatt T ClarkAsma R BankapurDaniel D'AndreaRosalba LeporeChristopher S FunkIndika KahandaKarin M VerspoorAsa Ben-HurDa Chen Emily KooDuncan Penfold-BrownDennis ShashaNoah YoungsRichard BonneauAlexandra LinSayed M E SahraeianPier Luigi MartelliGiuseppe ProfitiRita CasadioRenzhi CaoZhaolong ZhongJianlin ChengAdrian AltenhoffNives SkuncaChristophe DessimozTunca DoganKai HakalaSuwisa KaewphanFarrokh MehryaryTapio SalakoskiFilip GinterHai FangBen SmithersMatt OatesJulian GoughPetri TörönenPatrik KoskinenLiisa HolmChing-Tai ChenWen-Lian HsuKevin BrysonDomenico CozzettoFederico MinneciDavid T JonesSamuel ChapmanDukka BkcIshita K KhanDaisuke KiharaDan OferNadav RappoportAmos SternElena Cibrian-UhaltePaul DennyRebecca E FoulgerReija HietaDuncan LeggeRuth C LoveringMichele MagraneAnna N MelidoniPrudence Mutowo-MeullenetKlemens PichlerAleksandra ShypitsynaBiao LiPooya ZakeriSarah ElShalLéon-Charles TrancheventSayoni DasNatalie L DawsonDavid LeeJonathan G LeesIan SillitoePrajwal BhatTamás NepuszAlfonso E RomeroRajkumar SasidharanHaixuan YangAlberto PaccanaroJesse GillisAdriana E Sedeño-CortésPaul PavlidisShou FengJuan M CejuelaTatyana GoldbergTobias HampLothar RichterAsaf SalamovToni GabaldonMarina Marcet-HoubenFran SupekQingtian GongWei NingYuanpeng ZhouWeidong TianMarco FaldaPaolo FontanaEnrico LavezzoStefano ToppoCarlo FerrariManuel GiolloDamiano PiovesanSilvio C E TosattoAngela Del PozoJosé M FernándezPaolo MaiettaAlfonso ValenciaMichael L TressAlfredo BensoStefano Di CarloGianfranco PolitanoAlessandro SavinoHafeez Ur RehmanMatteo ReMarco MesitiGiorgio ValentiniJoachim W BargstenAalt D J van DijkBranislava GemovicSanja GlisicVladmir PerovicVeljko VeljkovicNevena VeljkovicDanillo C Almeida-E-SilvaRicardo Z N VencioMalvika SharanJörg VogelLakesh KansakarShanshan ZhangSlobodan VuceticZheng WangMichael J E SternbergMark N WassRachael P HuntleyMaria J MartinClaire O'DonovanPeter N RobinsonYves MoreauAnna TramontanoPatricia C BabbittSteven E BrennerMichal LinialChristine A OrengoBurkhard RostCasey S GreeneSean D MooneyIddo FriedbergPredrag RadivojacPublished in: Genome biology (2016)
The top-performing methods in CAFA2 outperformed those from CAFA1. This increased accuracy can be attributed to a combination of the growing number of experimental annotations and improved methods for function prediction. The assessment also revealed that the definition of top-performing algorithms is ontology specific, that different performance metrics can be used to probe the nature of accurate predictions, and the relative diversity of predictions in the biological process and human phenotype ontologies. While there was methodological improvement between CAFA1 and CAFA2, the interpretation of results and usefulness of individual methods remain context-dependent.