Are there any restrictions on species like we should include only species from same family or same order in the scientific classification?
Once we BLAST our sequence in NCBI, we get some similar sequences from different species. Is there any restriction to include them like identity % and query cover should be more and all?
Can we construct a tree by just selecting the sequence of a particular gene, in different species, even though they are not of equal length and with less similarity?
Dear Prithvi,
According to literature u can select homologous gene sequence for phylogenetic analysis. BLAST tool some times generate false positive result. u can concentrate on true positive result. more than 25% protein sequence identity and above 60% nucleotide sequence identity commanly considered for phylogenetic analysis. bUt it has some exceptions i.e. distantly related gene/protein sequences.
It defines only relationship between orthologous gene sequences. such type of alignments rarely contains a informative sites.
All depends on what are your questions. What is the motivation for reconstructing your phylogenetic tree or what do you expect to find or to explain?
Depending on this your tree might include sequences even from different domains (e.g. http://en.wikipedia.org/wiki/Tree_of_life_%28biology%29). On the other hand, if you are analyzing microdiversity within a species population your tree might include multiple sequences from strains from a same species (>=97% sequence identity in ssu_rRNA gene).
In general, homologous sequences is the requisite for a phylogenetic tree. Exactly same size of the sequences is not necessary because that is not is not expected to occur naturally anyway. However, the larger the shared homologous region of the sequences, the better the phylogenetic analyses are expected to behave because of some statistical assumptions behind them.
http://en.wikipedia.org/wiki/Tree_of_life_%28biology%29
I guess the best thing to it - just try it.
Depending on your sampling the usable and interesting genes for phylogenetic analyses can differ a lot.
Like Salvador said you might use some conservative regions for sampling sets across a broad range of taxa, but for population studies you need probably several single copy genes.
My adivse: Get some sequences from the respective gene from GenBank, align them and then do a quick and dirty maximum parsimony or likelihood analysis just to check if your selected region has enough information to give you a supported phylogeny
- Badges
- Science topic