Unannotated sequence

[ikim]

Hello all,

I have a mystery sequence that I was hoping y'all could help me with.
I've assembled a transcriptome set from both illumina and 454 datasets of the same plant sample. My approach to functional annotation has been to use blast searches against nr, refseq and hmmer searches vs interpro.
After performing expression analysis, I found that my 3rd highest expressed contig a long with a couple more very highly expressed ones were unannotated.
This 1200 bp sequence is found near identical in both my illumina and 454 assemblies. It has no hits vs nt, nr, refseq, cdd, interpro... it also seems to exist as the linking part of an assembly chimeric of 2600bp
(kinase-unannotated mystery seq-transferase).
Could anyone suggest ideas? I'm at a loss. Thanks very much!!

>c1
gttttatattagaattgacaaaaacataataataaaaaggttgtgtaactaaagatggcacttattgaaacaacattggctccgaatattactaaccataaccacaaccacggcgtttaagtggtgaacacagtataggtagaaacaaaatccataacatagcactaggaaaccctagaaaaacagggagacagagatgatgatcctctcctctccaaaagcattaccatccacactccagtcaccatcggtgctcaaaattattttactttctttcwsttaattgcctttgttcatcctcaactctctctttcctcccttttcttgcaattcaaattagtattccaatggccactgctgaggttgtatctgcagcgactgcattgcaagagaaagacacaactcatgaggaattgaacaagagcccagttgttgatgagaccaaggaagagaagccaacagaagaagtggtgacaccaccacccacatcagaagaggtcaaggaagacaaggctgatgcttcaatagaggaaccagtagccactacagatcaagcagaagccactgctgaagaagagaaagcagaggaggcacaagttgaggaggtaaaggaaacaaaggattcagttgaagaggagaaagcagtggttgaggagactaaagaagaggaatcaaaagaagataaggttagtactcctgaaccagtagcacctgaagagaagacccacgaaactacacctactactactaaagatgttagtgagagtactgttgaagcagaagagaaagttgttcaatcagagataccagttgaggaagccaaagcaacagaagcagaagagaaagttgttgccgcagagactactccagtagttgagaaggctgaggagtagatttgatctgtgttccccaggattctgattgtctggccagctagtagtgctggatgtttgtgtagtgggtgttttatatagaagcttgcatgttaagagttgatgagttttaatgtagtaaaagaatctatgtaggttatgaggcttggaagttagtatattaacttggtattttcttgggcccaccaagtaccattctgccatgtggcaaaggtccaatgccccaaaaatatgtttcaaattcccaaagcttttgtttggtggagggacactctctctattgcagtgtggagggttgtgtactggttggtctcgtgtactgtggatvdwcatsrsagttgcagcwacatyyraagccaaagagtaatcacagattttaaaagtaagtgttgggccctg

[Jeremy]

A functional but non-protein coding RNA? Maybe something in the ENCODE database could help you.

[ikim]

Thanks for your suggestion Jeremy. I'll inquire with ENCODE.
I would have thought that non coding rna would have nt annotations at least.

[rbs.sachin]

Hello all,

I have some 454 data and in that there are approx 138 sequences that I am not able to found in Mouse Genome. the are unplaced or uncharacterized.

So, Please help me to find them.??????//

[kmcarr]

Hello all,

I have a mystery sequence that I was hoping y'all could help me with.
I've assembled a transcriptome set from both illumina and 454 datasets of the same plant sample. My approach to functional annotation has been to use blast searches against nr, refseq and hmmer searches vs interpro.
After performing expression analysis, I found that my 3rd highest expressed contig a long with a couple more very highly expressed ones were unannotated.
This 1200 bp sequence is found near identical in both my illumina and 454 assemblies. It has no hits vs nt, nr, refseq, cdd, interpro... it also seems to exist as the linking part of an assembly chimeric of 2600bp
(kinase-unannotated mystery seq-transferase).
Could anyone suggest ideas? I'm at a loss. Thanks very much!!

>c1
gttttatattagaattgacaaaaacataataataaaaaggttgtgtaactaaagatggcacttattgaaacaacattggctccgaatattactaaccataaccacaaccacggcgtttaagtggtgaacacagtataggtagaaacaaaatccataacatagcactaggaaaccctagaaaaacagggagacagagatgatgatcctctcctctccaaaagcattaccatccacactccagtcaccatcggtgctcaaaattattttactttctttcwsttaattgcctttgttcatcctcaactctctctttcctcccttttcttgcaattcaaattagtattccaatggccactgctgaggttgtatctgcagcgactgcattgcaagagaaagacacaactcatgaggaattgaacaagagcccagttgttgatgagaccaaggaagagaagccaacagaagaagtggtgacaccaccacccacatcagaagaggtcaaggaagacaaggctgatgcttcaatagaggaaccagtagccactacagatcaagcagaagccactgctgaagaagagaaagcagaggaggcacaagttgaggaggtaaaggaaacaaaggattcagttgaagaggagaaagcagtggttgaggagactaaagaagaggaatcaaaagaagataaggttagtactcctgaaccagtagcacctgaagagaagacccacgaaactacacctactactactaaagatgttagtgagagtactgttgaagcagaagagaaagttgttcaatcagagataccagttgaggaagccaaagcaacagaagcagaagagaaagttgttgccgcagagactactccagtagttgagaaggctgaggagtagatttgatctgtgttccccaggattctgattgtctggccagctagtagtgctggatgtttgtgtagtgggtgttttatatagaagcttgcatgttaagagttgatgagttttaatgtagtaaaagaatctatgtaggttatgaggcttggaagttagtatattaacttggtattttcttgggcccaccaagtaccattctgccatgtggcaaaggtccaatgccccaaaaatatgtttcaaattcccaaagcttttgtttggtggagggacactctctctattgcagtgtggagggttgtgtactggttggtctcgtgtactgtggatvdwcatsrsagttgcagcwacatyyraagccaaagagtaatcacagattttaaaagtaagtgttgggccctg

By eye one can see that this sequence very low complexity (repetitive). Did you run your BLAST search with the default parameters? The default parameters for BLAST mask low complexity regions prior to the search to speed it up by avoiding low information hsps. Using the low complexity filter for this sequence is means very little of the sequence will actually be searched.

Run a BLASTX of this sequence against the NCBI nr protein database with the low complexity filter disabled. You will see a very large number of hits to proteins in nr. Now most of those are annotated as hypothetical or predicted but at least it's a place to start.

[ikim]

Thanks very much for your advice Kmcarr and bringing up the low complexity issue. That makes a lot of sense, DUST or Seg was on during my initial searches.
I'd run repeatmasker on this sequence though with settings to determine low complexity regions but it did not report anything. And unfortunately, turning off filtering on the NCBI blast server did not return any additional results for me.

[kmcarr]

And unfortunately, turning off filtering on the NCBI blast server did not return any additional results for me.

Sorry, forgot to mention that you also need to disable Composition Based Statistics. If you are using the NCBI web BLASTX select "No adjustment" from the pop-up for Compositional adjustments. If you are running blastx on the command line use "-comp_based_stats 0" (that's a zero, not the letter 'O').

Disabling low complexity filtering and composition based statistics your sequence returned > 100 significant hits (E ≤ 1e-10) vs. the nr protein database. What is being matched is a general pattern of very glutamate rich proteins.