This page last modified: Feb 19 2007
keywords:gene,space,sequence,protein,gss,peptide,amino,acid,blast,search,alignment,cdna,strand
description:How to verify translation between Blastx results and an original gene space sequence.
title:GSS search and protein aligment explained for programmers.
This page is intended to scroll horizontally in your web browser. This
obviates goofy line wrapping conventions that are irritating in
multi-line alignments.
BlastX translates input gene sequences into all 6 reading frame, then
translates each sequence into a protein. The translated proteins are
searched against a database of proteins (usually proteins of known
identity). This process also applies when comparing two libraries of
gene sequences. Each library is translated into the 6 reading frames
and searched with Blast (TBlastX, resulting in 12 searches). Genes vs
proteins is 6 searches, genes vs genes is 12 searches. Clearly it
would be computationally simpler to search genes vs genes without
translation, however, gene sequences are not well conserved in nature,
whereas protein sequences are reasonably well conserved. The extra
computational effort makes biological sense.
Below are the details of a gene sequence translated to a protein, and
matched up with a protein hit. Sequence numbering starts with one, not
zero. Sequence from and to are inclusive numbers.
Reading frames are 3, 2, 1, -1, -2, and -3.
Forward reading is 5' to 3' (five prime to three prime). Reverse
strands are 3' to 5'. For reasons not entirely clear to me,
translation is often done on the complementary strand. Complement is
the Perl tr// operation:
$seq =~ tr/acgtACGT/tgcaTGCA/;
Uppercase and lowercase are often not well defined, and seem to
reflect the whim of the biologist or programmer involved. When
defined, typically uppercase are better quality values (such as Phrap
base-calling quality values). In the example below, case has little
meaning.
This is not the best hit for bs_fk 82842, just one I chose at random
to test. The best hit (based on e score) is for a longer sequence
(albeit it has a gap), and is in reading frame -1 instead of the
reading frame -2 shown below.
My database br_pk=149358
My database bs_fk=82842
Here are a few relelvant fields from the blast results:
hsp_hit_from | hsp_hit_to | hsp_positive | hsp_query_frame | hsp_query_from | hsp_query_to
--------------+------------+--------------+-----------------+----------------+--------------
147 | 187 | 36 | -2 | 474 | 596
hsp_qseq | hsp_midline | hsp_hseq
-------------------------------------------+-------------------------------------------+-------------------------------------------
VGVPFERIKRLGEEDNFWTSGVTGPCGPCSEIYYDFHPERG | +G+ +RI +LGE+DNFW+SG TGPCGPCSE+Y+DF PE+G | IGIHPKRIIKLGEKDNFWSSGKTGPCGPCSELYFDFKPEKG
Original GSS from the cowpea database. "|" marks a base location. Numbering starts at 1.
AAAACCATATGAAACATTTGCTAACTTAGAGGCCTCTTCTATGATGGGGAAAATCAAGTCAGTTTCATAATTGTTTGGAACCTGCAGAAAGCATTGGCTATATGTAAGATGGACTTTAGGAGCATGTGAAATGATAGTTCTAGTGCTCCAGAAGTTCTTCCTAAAATGTAAACGTAACAACTACAACAACAAAAAATCAACATAGTGAACATATACCTGTTGAAGAATACGAGCCATACGCTCCAGTCCCAGGCCAGTGTCTATGTTCTTCTGTTTTAGAGGTTCAAGAGAACCATCATCCTTTTTGTTGTATTGCATGAAGACTAAGTTGTAGAACTCTATAAACCTTGTATCGTCATTGAGATCCTGCTTTCATCCCCCAGGTCCCAACCACCCAACAAATATAAAATATTGTCTCTGAAATGATATTGAATCCTTCAATCTAAAAATGGAATACTTACAGCATCTACATATCCCCTCTCAGGATGAAAATCATAATAAATTTCTGAGCAAGGACCACAGGGACCTGTCACTCCACTAGTCCAGAAGTTGTCTTCTTCTCCCAACCTTTTTATACGTTCAAATGGGACACCAACCTGCACGGAAATCAAGTAAGCAATAGGAACAATAATCCAATACTCTTTGTAAAATGTTCACTAGTCAAAAcaGaagtTAGTtatGTAAGCAAaTAATAcg
|1 |474 |596
474 thru 596 3'5' frame -1 (this sub-nucleotide is exact in frame -1, even though the whole GSS is read in frame -2).
(translated seq is reverse, complement)
gttggtgtcccatttgaacgtataaaaaggttgggagaagaagacaacttctggactagtggagtgacaggtccctgtggtccttgctcagaaatttattatgattttcatcctgagagggga (translated seq)
V G V P F E R I K R L G E E D N F W T S G V T G P C G P C S E I Y Y D F H P E R G
tcccctctcaggatgaaaatcataataaatttctgagcaaggaccacagggacctgtcactccactagtccagaagttgtcttcttctcccaacctttttatacgttcaaatgggacaccaac (original seq)
Check peptide (from below):
V G V P F E R I K R L G E E D N F W T S G V T G P C G P C S E I Y Y D F H P E R G
Gene sequence translated to protein sequence (peptide) by: http://ca.expasy.org/tools/dna.html
Reading frame -2 (negative frame is 3' to 5')
The protein squence displayed here has been "fixed" with two translations, and spaces removed:
$ = Stop
^ = M
VLFAYITNFCFD$$TFYKEYWIIVPIAYLISVQVGVPFERIKRLGEEDNFWTSGVTGPCGPCSEIYYDFHPERGYVDAVSIPFLD$RIQYHFRDNILYLLGGWDLGDESRIS^TIQGL$SSTT$SSCNTTKR^^VLLNL$NRRT$TLAWDWSVWLVFFNRY^FT^LIFCCCSCYVYILGRTSGALELSFH^LLKSILHIANAFCRFQTI^KLT$FSPS$KRPLS$Q^FH^V
VGVPFERIKRLGEEDNFWTSGVTGPCGPCSEIYYDFHPERG (query peptide, hsp_qseq)
+G+ +RI +LGE+DNFW+SG TGPCGPCSE+Y+DF PE+G (midline, hsp_midline)
IGIHPKRIIKLGEKDNFWSSGKTGPCGPCSELYFDFKPEKG (hit peptide sequence, hsp_hseq)
MKSQTKNTPITGDEIRKEFLNFYHEKLHKIIPSASLIPDDPTVMLTIAGMLPFKPVFLGLKERPSKRATSSQKCIRTNDIENVGVTARHHTFFEMLGNFSFGDYFKKEAIEWAWELVTDIYGLSAENIIVSVFHEDDDSVKIWKEDIGIHPKRIIKLGEKDNFWSSGKTGPCGPCSELYFDFKPEKGVQNIDLEDGDRFIEFYNLVFMQYNRDPDGQLTDLKYKNIDTGMGLERMAQILQKKKNNYETDLIFPIIQKASEISKIDYYSSGERTKISLKIIGDHIRAVIHLISDGVIASNLGRGYILRRLIRRMVRHGRLLGLKNEFLSKLASVGIKLMQENYPDLKNNCDHILSEIKIEEIRFRETLERGEKLLDELISSGQKMITGFKAFELYDTYGFPLELTEEIAQENNIGVDVKGFDKEMSAQKERAKAASQIIDLTLEGSLEREIDLFDKTLFNGYDSLDSDAEIKGIFLESTLVKQASEGQKVLIVLDQTSFYGESGGQVGDIGTILSNDLEVVVDNVIRKKNVFLHYGIVKKGILSLGQKVKTKVNDLARAKAAANHTATHLLQSALKVVVNESVGQKGSLVAFNKLRFDFNSSKPITKDQIFKVETLVNSWILENHSLNIKNMAKSEALERGAVAMFGEKYDDEVRVVDVPSVSMELCGGTHVKTTSELGCFKIISEEGISAGVRRIEALSGQSAFEYFSDKNSLVSQLCDLLKANPNQLLDRVNSLQSELINKNKEIQKMKDEIAYFKYSSLSSSANKVGLFSLIISQLDGLDGNSLQSAALDLTSKLGDKSVVILGGIPDKENRKLLFVVSFGEDLVKRGMHAGKLINDISRICSGGGGGKPNFAQAGAKDIDKLNDALEYARKDLRTKLHSYSDK
|1 |147 |187