Frame disagreement between browser and Blastx alignments

Ask questions about annotation of D. erecta, D. mojavensis, and D. grimshawi projects here.
Post Reply
nkokan
Posts: 10
Joined: Fri Jun 06, 2008 5:24 pm
Location: Cardinal Stritch University

Frame disagreement between browser and Blastx alignments

Post by nkokan » Wed Apr 09, 2014 10:57 pm

We are working with Dbiarmipes dot_Aug2013_contig1, feature 1.1 (PlexB ortholog). Blastx alignments gave the following predicted coordinates, 13042-12542 in frame -3 for the exon. In the browser, we found stop codons for the exon in this region in frame -3. However, there is a start codon at 13042 but it is in -2 frame. How should this be resolved? Should this issue be reported at the time the project is submitted?
Nighat P Kokan, PhD
Associate Professor
Department of Natural Sciences
Cardinal Stritch University
Milwaukee, WI 53217
414.410.4138

wleung
Posts: 185
Joined: Sun Feb 04, 2007 7:41 pm
Location: Washington University in St. Louis

Re: Frame disagreement between browser and Blastx alignments

Post by wleung » Thu Apr 10, 2014 12:57 am

Unfortunately, I could not reproduce the discrepancy. The blastx search of the CDS plexB:6_1762_0 against contig1 shows that the match is in frame -2, consistent with the results shown on the GEP UCSC Genome Browser:
blx_plexB_6_17862_0_contig1.png
blx_plexB_6_17862_0_contig1.png (61.87 KiB) Viewed 6298 times
One possible explanation for the discrepancy is that the contig1 sequence you used in the blastx search might be different than the sequence shown on the genome browser. In addition to accessing the contig sequence through the "DNA" functionality of the UCSC Genome Browser, the sequence is also available in the src folder in the annotation package. You can download the annotation package through the GEP Data Repository. For your convenience, you can also download the package through the following link: dbiarmipes_dot_Aug2013_contig1

Please re-run the search using the contig sequence in the package and let me know if the problem persists. Thanks.

nkokan
Posts: 10
Joined: Fri Jun 06, 2008 5:24 pm
Location: Cardinal Stritch University

Re: Frame disagreement between browser and Blastx alignments

Post by nkokan » Thu Apr 10, 2014 2:37 am

That is great to see a -2 frame but I still have to resolve the -3 frame issue as she ran these twice. We used DNA from the Gander Browser "get DNA" feature for the contig. My student was unable to open the sequence file in the src folder. I will re run the blastx with the project sequence file and the "get DNA" feature. Should she be using the masked or unmasked sequence? Thanks for resolving this issue to some extent. I will ask her to re run the blastx one more time and report back.
Thank you!
Nighat P Kokan, PhD
Associate Professor
Department of Natural Sciences
Cardinal Stritch University
Milwaukee, WI 53217
414.410.4138

wleung
Posts: 185
Joined: Sun Feb 04, 2007 7:41 pm
Location: Washington University in St. Louis

Re: Frame disagreement between browser and Blastx alignments

Post by wleung » Thu Apr 10, 2014 4:26 am

> My student was unable to open the sequence file in the src folder.

The sequence file in the "src" folder is a plain text file so you can open the file in any text editor (e.g. WordPad on Windows and TextEdit on the Mac). You may need to manually select the program if the operating system does not recognize the .fasta extension.

Alternatively, you can select the fasta file using the "Browse" button without opening the file:
ncbi_blast_browse_button.png
ncbi_blast_browse_button.png (71 KiB) Viewed 6294 times

> Should she be using the masked or unmasked sequence?

In general, I would recommend that you should use the unmasked sequence when you use BLASTX or TBLASTN to map each exon to your contig sequence. This is because there is a possibility that an exon might have been incorrectly masked by RepeatMasker (e.g. because of low complexity or simple repeat sequences). I would use the masked sequence when performing a database search (e.g. against the RefSeq or nr databases) to reduce the number of spurious matches (e.g. to transposons).

nkokan
Posts: 10
Joined: Fri Jun 06, 2008 5:24 pm
Location: Cardinal Stritch University

Re: Frame disagreement between browser and Blastx alignments

Post by nkokan » Thu Apr 10, 2014 6:07 am

I repeated the blastx with compositional matrix "no adjustment" and filters turned off and I got an alignment in -2 frame with the same predicted coordinates that you posted earlier. I think the issue with -3 frame that the student had might have been due to the parameters being not set correctly. I think the problem is solved for now. Thank you for your help.
Nighat P Kokan, PhD
Associate Professor
Department of Natural Sciences
Cardinal Stritch University
Milwaukee, WI 53217
414.410.4138

Post Reply