From lionel.guy at icm.uu.se Tue Dec 10 14:41:22 2013 From: lionel.guy at icm.uu.se (Lionel Guy) Date: Tue, 10 Dec 2013 14:41:22 +0100 Subject: [genoPlotR-help] genoPlotR help In-Reply-To: <9601DD6757D2244CADC3B9007A4639A443C6A88F@exchange2-1> References: <9601DD6757D2244CADC3B9007A4639A443C6A88F@exchange2-1> Message-ID: <78AA6025-39B8-4B3D-9180-B835C7F1C9E8@icm.uu.se> Hi Adnan, That's because blastp doesn't use genomic coordinates: the numbers in the hits refer to positions inside the proteins, not genomic positions. You can either convert protein positions into genomic positions (which is a pain in the ass), or use tblastx. If you use tblastx, beware that it had, in previous versions at least, a hard-coded limit of 400 hits returned for a pair of sequences (which is the case when you match two genomes against each other). NCBI might have fixed that in newer versions. If it is not the case, you'll miss most of the hits. The solution is to break down the jobs in smaller parts, for example in using the -subject_loc option. HTH, Lionel On 10 Dec 2013, at 13:16 , Adnan Niazi wrote: > Hej Lionel, > > I am using genoplotR for drawing some genome maps. I tried one by doing blastn of 10 genomes using genbank which worked perfectly. But when I make one with blastp and using *.faa and *.ptt files I get only blue blocks (CDS) on the top of each map but not the connecting lines between the two genomes. Can you please tell how can I show lines in the map. > > > > regards, > adnan -- Lionel Guy Molecular Evolution, Uppsala University, Uppsala, Sweden postal address: Box 596, SE-751 24 Uppsala; visiting address: BMC B7:213e, Husargatan 3, SE-752 37 Uppsala phone: +46 18 471 6129, mobile +46 73 976 0618 lionel.guy at icm.uu.se