Bold face: keyword "evaluate prediction accuracy"
NOTE: input MUST be in column format and MUST contain prediction and experimental assignment.
Effect: scores for per-residue and per-segment accuracy are returned
Joe Sequencer, Department of Advanced Protein Research, National Univeristy, Timbuktu joe@amino.churn.edu evaluate prediction accuracy # COLUMN format , incredulase from paracoccus dementiae, translated from cDNA NAME AA PSEC OSEC first M L L first Q L L first T L H first S H H first E H H first R H H first L H H first A L H first G L L first V L L first K L L first Q H H first Q H H first S H H first I H H second G L L second V L L second K E L second Q E H second Q E H second S L H second I L H
Compulsory information: (1) predicted secondary structure (PSEC) in either of the states H=helix, E=strand, or L=rest; (2) observed secondary structure (OSEC) in either of the states H=helix, E=strand, or L=rest (e.g. from DSSP assignment); (3) if more than one protein is used, simply append all requested proteins (in that case make sure that the first column (NAME) lists a unique protein name).
Optional: (1) sequence (AA) in one-letter code; (2) name of protein (compulsory for more than one protein).