Entering edit mode
10.7 years ago
chevivien
▴
90
Hallo...how to one examine whether homology method used is feasible for identification of a particular gene family e.g HSP. Suppose i have identified 550 sequences suspected to be a particular gene family and i want to know if actually this are my target genes which approach will suitable?. Given that some previous research has been done before of this family and around 200 was identified then...Will reciprocal BLAST work well?if yes how will i know for instance that some of my identified genes are similar with the genes identified earlier on by other researchers given that the Accesion numbers of my identified sequences are different with the ones identified earlier on?
for HMMER you may want to do statistical validation, for example, confusion matrix( calculating true positives, false positives, true negatives, false negatives) then, sensitivity, specificity, precision, accuracy and Mattew's correlation coefficient, if I understood well your question
Let me paraphrase my Q....supposing i have identified a set of sequences of a particular gene family using a combination of BLAST and HMMER and i want to find out whether the sequences identified are actually for target gene family. From my literature search i understand you can compare the identified sequences with ones identified earlier on by other researchers...if you find that some percentage of my identified seq are similar to those identified earlier it implies what i got might be potential gene family iam interested in.. Part of this genes has been identified before by other researches and say they obtained 200 sequences but during my homology search i got 500.. To confirm if my identified sequences are the target genes i might want to compare and find out if any of my sequences are similar to ones identified by other researchers... My problem is that if iam to compare how will i go about given that the ID numbers of my sequences are different from ones identified before..e.g mine starts with something like lcl|Os01t0359600-00 but for earlier research Id starts with gi|380777041|gb|AFE62091.1|(NCBI ACCESSION).Will reciprocal BLAST on this case? Or rather what is the best way to validate homology search results?