《bryantandaltschul1995》.pdf

! Statistics of sequence-structure threading Stephen H Bryant and Stephen F Altschul National Institutes of Health, Bethesda, USA Thepast two yearshaveseenthe rapid development of newrecognition methodsfor protein structureprediction. These algorithms‘thread‘ the sequence of one proteinthrough the known structure of another, looking for an alignment that corresponds to an energetically favorable model structure. Because theyare based on energycalculation,ratherthanevolutionary distance,thesemethodsextendthe possibility of structureprediction by comparative modelingto a lawr class of new sequences, where similarity to known structures is recognizable by no other means. The strength of the evidence they offer should be judged by objectivestatistical tests, however, so as to ruleoutthepossibilitythatfavorable scores arisefromchance factorssuch as similarityof length,composition,or the consideration of a large number of alternative alignments. Calculation of objective pvalues by analytical meansis not yet possible, but itwould appear that approximate values maybeobtained by simulation,as they arein gapped, global sequence alignment.Weproposethattheresults of threadingexperimentsshould