Bioinformatics - protocols  B

---

Basic sequence manipulation

>cDNA_B_At
CCGCGGTGGCGGCCGCTCTAGAACTAGTGGATCCCCCGGGCTGCAGGATC
CCGGTGACCCCGGCAAAGCTTGCTTAATCCGAAGACGTTTCTGTTTCATC
TTCTTAAATCCGGGCCAACNGCGTTTACGAGACTAAACGCGTTTCTCTTT
AGGGCTTAATTATTATCCAGAGATGGCTCATCATAGCAAATGTTTACAAA
CGTTGGATTTAGCTTGTAAAGAGCTGAGATCTCGTGGCTTGTTTGTGAAG
CTTTTGGAGGCAATACTTAAAGCTGGAAACAGAATGAACGCGGGTACCGC
GAGAGGAAACGCTCAAGCGTTTAATCTAACCGCGCTTTTGAAGCTTTCGG
ATGTTAAAAGCGTTGATGGGAAGACTTCTTTGCTTAACTTTGTAGTGGAG
GAAGTTGTTAGATCGGAAGGAAAACGTTGTGTTATGAATAGAAGAAGCCA
TAGCTTAACACGAAGCGGTAGTAGTAACTACAATGGTGGTAATAGTAGTC
TTCAGGTTATGTCGAAAGAAGAGCAAGAGAAAGAGTACTTGAAGCTTGGT
TTACCAGTTGTTGGTGGATTGAGCTCTGAGTTTTCAAACGTGAAGAAAGC
TGCTTGTGTGGACTATGAAACGGTTGTTGCAACTTGTTCTGCTCTTGCGG
TTAGAGCGAAAGATGCGAAAACGGTGATTGGAGAATGTGAAGATGGAGAA
GGAGGGAGGTTTGTGAAAACGATGATGACGTTTCNTGATTCGGTAGAGGA
AGAGGTGAAAATAGCGAAAGGTGAAGAGAGGAAAGTGATCCCTGA

• Look at the coding capacity of your cDNA using either the BCM sequence utilities 6 Frame Translation function, or the ExPASy Translation Tool . Make sure that you are NOT using the default "Verbose" option on ExPASy (or just try it once so that you know why avoiding it :-). (You might want  to look also at other translation tools such as The Protein Machine).
• Select the longest reading frame in the sense orientation and keep it as a FASTA file. (NOTE: you might not necessarily get a reasonable ORF - assume that this was a raw first run sequence, containing possible frameshifts).

Sequence similarity searches and domain structure analysis

• Use the protein sequence from the previous excercise as a query for a standard protein BLAST search of the non-redundant NCBI database. Uncheck the "Do CD Search option", otherwise keep default parameters. If you were unable to perform the search for server-dependent reasons, go to the SOS page. You should be able to produce a working hypothesis about the possible function of your cDNA already at this stage!
• From the BLAST results page, retrieve the protein sequence corresponding to the best hit (by clicking the link to the left from the gene description- see Figure, which is accidentaly from a nucleotide BLAST but the file structure is the same). If the BLAST server is down, go to the SOS page.


• Retrieve the corresponding protein sequence in FASTA format (or, in the worst case, from the SOS page). Run a protein BLAST search as above using the retrieved protein sequence as query and all default parameters including the CD-search.
• On the 1st screen (before "Format Results") you will see a summary of conserved domains found in your sequence by CD-search. KEEP THIS PAGE OPEN FOR FUTURE USE. If you lose it, you have to redo the search. If the server is down, go to the SOS page.
• On the BLAST results page, look up which kind of things the search has picked by looking at some of the alignments (links to the right from sequence description). Note also the number of entries corresponding to all hits with E-values better (i.e. lower) than 10^-4 (denoted as e-04) and repeat the BLAST search, disabling this time the low complexity filter (you can uncheck CD-search as well for increased performance). Again, look at the results and note the number of hits with E lower than 10^-4. Why is filtering the query a good idea? (Both BLAST results can be found on the SOS page).
• Perform a search for transmembrane or signal peptide sequences in your hypothetical protein using SignalP  and/or TMHMM.

Construction and interpretation of a protein sequence alignment

• On the resulting page (backup on SOS) chose "add query to the alignment", keeping all other parameters default, and examine the output (could be found also on the SOS page).Note how the identical sequence picked from the database was (mis)aligned.
• Repeat the previous task using the "most diverse set" and "top of CD search" options.

• Open MACAW on your computer (should be OK over the browser). Go File... New Project..., select Sequence type= protein and Significance=number. Chose the PAM120 scoring matrix and import sequence (Sequence... Import...) from the file you just created. Save the Macaw file on your computer before proceeding further in order to decrease the likelihood of crash.
• Examine the performance of Gibbs sampler vs. Segment pair overlap methods and construct an alignment of the protein sequences.
• (Optional): repeat the last task using a different scoring matrix.

Gene building: searching for coding sequences in chromosomal DNA

>locusB
cccctataaaaagtattaaaaaggactgatacaataatgtatataaatat
cctaaaagatcttaattttgtaaatttattgttgtatattctaaacccgc
aatattagaatgatgatttagtaaacaagaaagacaaaataaataattaa
ttttagctagaaaagatgaaataaacactcatgatttaagccatacaaat
cgaagccccttgggttcagcatttctcaccaagtaaataccatcacctct
ggaaacccatttacgtacttgaccacatcttttattagcggctcctctgt
atgctctccatatgttatacacactatgatgccttaagatttattcacga
cgatttaatcagatacgcttatggattgccaaagatgatgccatctactt
agagaaaaacaatggaaagcgagaacgcatgtataattggaataaaaatt
aatatggttttcatatatctaaaaaattggacatttgaagccttaataaa
ttatactatgtaaaaatacttgtttatgaatgtaaattataataaattac
gatttaattagggaaatattgactatatatttcacccaaatattgaatgt
aaattttattttccaatacttttgcacatttaagaaattttcggatgtat
ttcctaaagaatattaccttttttgttttttaaaccatgcctttttgttt
tacacgttcataaatgcatgttccatacgcattaccataatttaatttga
acttaattttctctaggaatggtgatgatccactaccactatcattgatt
tcattccatattcctttgaccgactgaaattacgttggaaatagtatatt
ttgatgaataatttatttactcggaaaaaagaggtcaagttattaatagt
aagtacatatacattatcaattaagaattcaattgagttttaaggaaaat
cctattaatttgtttggtattcggtatttgttagttctaaggaattgaat
ttcccgattatacatcattataacgttctcaagttccaaacttgcaaccc
acattttgtcgatattctcaaatgtgaattcattcaatttcccatagaaa
acataaatttgcacttaaagttaacaattgaaatcgtatctaaatgggaa
tgtttttggcttttagtgttagacttccaaagcgtcaaaaatatttctag
aaagagcacaaaaaataagcaacgccactacttttggacaaagtcaacga
taacacacatcaaccgcaccagctccataaaagtccatctcacgaaaacg
attctagtcaaactacctaaaacacccttatatttacatacaacccaatc
ccactaacaagggtattttcgtcaatcacaaaatttatcaccgacccggg
aagaagaagaagaacagatcaactaatttctgctttcaactccacattaa
accaaaacctccaaaaagaatcatttatttaaattatcttcccgttttaa
gttcctgagatttttgggaattgtaaatttgaagaaaattaaacaaagac
gtgttttcatttttttttttgtttcctttattgatctctctctatctctc
taaatgagctaaatcgttaatggctgccatgtttaatcatccatggccta
atttaaccctaatttacttcttcttcatcgtcgttttaccattccaatca
ctttctcaatttgattctcctcaaaatatcgaaactttcttccccatctc
ttcactctcccctgttccaccaccgcttcttccaccttcgtcaaacccat
ctccgccgtcgaataattcatcatcttcggataaaaaaacaatcaccaaa
gctgtccttataacagcagcaagtactttacttgtagctggagttttctt
cttctgcctccaaagatgtatcatcgcacggagacggagagacagagttg
gaccagtcagagtcgaaaacactttacctccgtatcctcctcctccgatg
acgtcggcggcggtgactacgactactttggctagagaaggattcacgag
gtttggtggtgtgaaaggtttgattcttgatgagaatggtcttgatgtgt
tgtattggagaaagctacagagtcagagagaaagaagtgggagtttcagg
aaacagatcgtcaccggagaagaagaagacgagaaagaagttatttatta
caagaacaagaagaaaacagagcccgttacagagattcctcttcttagag
gaagatcatctacttctcacagtgttatccataacgaagatcatcagccg
ccaccgcaggtgaaacagagtgaaccaacaccaccaccgccaccaccgtc
aattgcggtgaaacagagtgcaccaacgccatcgccacctcctccgatta
agaagggttcttcaccatcgccaccgccacctccaccggtgaaaaaggtt
ggagctttatcatcatcagcttcgaaaccaccacctgcgccggttagagg
agcaagtggaggagagacttcgaaacaagtaaagttgaagcctttacatt
gggataaagtaaaccctgattccgatcattcaatggtttgggacaaaatc
gatcgtggatcattcaggtatatatttatttcgaaagttagggcttttgc
ttcaatcaattgaaaaaaccctaatttgtttttgtttcttctcagtttcg
atggcgatttaatggaagctctgtttggatacgttgccgtggggaagaaa
tcaccagaacaaggcgatgagaaaaaccctaaatcaacgcaaatattcat
acttgatccgagaaagtctcaaaacacagcgattgtgctcaaatcattag
gtatgacacgtgaagagcttgttgaatcactcatagaaggaaacgatttc
gtgccagacactcttgagaggttagctagaatagctccaacgaaagaaga
acaatcagccattcttgaattcgacggtgacacggcaaagcttgctgatg
cggagacgtttctgtttcatcttcttaaatccgtgccaaccgcgtttacg
agactaaacgcgtttctctttagggctaattattatccagagatggctca
tcatagcaaatgtttacaaacgttggatttagcttgtaaagagctgagat
ctcgtggcttgtttgtgaagcttttggaggcaatacttaaagctggaaac
agaatgaacgcgggtaccgcgagaggaaacgctcaagcgtttaatctaac
cgcgcttttgaagctttcggatgttaaaagcgttgatgggaagacttctt
tgcttaactttgtagtggaggaagttgttagatcggaaggaaaacgttgt
gttatgaatagaagaagccatagcttaacacgaagcggtagtagtaacta
caatggtggtaatagtagtcttcaggttatgtcgaaagaagagcaagaga
aagagtacttgaagcttggtttaccagttgttggtggattgagctctgag
ttttcaaacgtgaagaaagctgcttgtgtggactatgaaacggttgttgc
aacttgttctgctcttgcggttagagcgaaagatgcgaaaacggtgattg
gagaatgtgaagatggagaaggagggaggtttgtgaaaacgatgatgacg
tttcttgattcggtagaggaagaggtgaaaatagcgaaaggtgaagagag
gaaagtgatggagcttgtgaaacgtacaacggattattatcaagcaggag
ctgttacaaaggggaagaatccacttcatttgtttgttatcgttagagat
tttcttgccatggttgataaagtttgcttagatattatgagaaatatgca
gaggaggaaggttggtagtccgatatcgccttcttcgcagcggaatgcgg
tgaaattcccggttttgcctccgaatttcatgtcggacagagcttggagt
gattctggtgggtcggattctgatatgtgagagtcaagatttgttatatg
taaatactaaatagtagaagcattttgggtattgattagcattgaaagat
gttgaattgtttatagatttatcagtccaaagcattggacttgagtataa
tttgttccttgtataaataaacaattttgctttaagacctttccatgttt
atgaacatgtcttctttaacttcacatagaccttttgtttacgtaagaac
taataatactaaattgtttgataattctaaatgtgaaagtgaaccactat
atagtgtgaacttggctttattgaattctttttaaaaaaatttctccaga
gctttagatgtaggagttaatattttcacctaacatagcctcttttttat
gtttctctatcaactaacactaaatttgtggatgaagactaaattaacat
aagtttatctattaactaacaacctaccagtttgatgcttgtaaatatga
aacttcaacgttataaagactatatggtgtgaactttttatccatcttta
ttgacttttaaaattttcttaatttgagtaaacaaaagcagaagcttttt
aaaggatgcaggagttgatttttgtatatgaacaaaacatatacttctcc
cttagacgaatttggagctatcattcttggtttcaaactttttaataatt
tgagctttaaagcaaaatggcaactttatattgatcactagtccacaaca
ctttctctgccttttcctcaatagcaacgcgtagtcaagaagaagaacgt
gtttaacatggaccaatcttgattaagataatagtatgatcaaatgctta
tataaacacactaaaaaggaatcaaatttaaccattccacaaatcaccaa
caaaatttaatgaatcatgtctctgcttctaaagatgttattattttcct
tattcttcttctatatggcttcaatttctcaatgctcagacccaaccggt
ggacagtttagcttcaacggttacttgtacaccgatggagttgcggatct
aaacccggacggtttgttcaaactcataacttcaaagaca

• Try to predict the exon-intron structure using the NetGene2 server (chose Arabidopsis, and take care to load only sequence, not the FASTA header, into the bottom box). (If server down, the results can be found on the  SOS page.)
• (Optional): try the same using the Web Gene Gene Builder interface. Take into account the results of the previous excercise (the simplest way is using the best protein match, provided for your convenience below, as key protein). Chose Direct strand, Gene Model, Sequence error report, Use EST mapping and Complete gene model; switch off the Protein homology search. Don`t forget to select organism = Arabidopsis, and the plant scoring matrix! (NOTE: there are currently problems at the server, but look at least at the interface to get an idea what it should do).
• Look up results of other prediction methods - GenScan and GeneFinder (use links or contact group A) and compare the results.