As we’ve seen, it’s tempting to use top hits from BLAST or FastA results to classify unknown proteins –but this may lead us (& especially computer programs) to false functional conclusions PSI-BLAST is more sensitive than BLAST, because it creates a profile from hits above a given threshold –but this too can cause problems –let’s take a closer look Misleading results from BLAST