Functional Genomics & Bioinformatics - progressive alignment question

[MelonBar]

Hey guys, I have a question about the progressive alignment method of multiple sequence alignment. I'm trying to understand how it is actually done, so I think you do pairwise matches for all protein sequences, write the most similar pair of sequences down, and then you: add subsequent individual sequences that are most similar to the first pair and so on?

I realize it's a specific question that only people who do the subject or have taken it can answer, but the lecturer never replies to my emails so here I am hoping for the best!

cheers

The Melonator

[scribble]

Pretty sure that's correct. You do the pair wise match between all of them and select the best match. Then you add the sequence that has the highest score with both of those to the alignment. Then you add the sequence that is most similar to the three and repeat til you're done.

[MelonBar]

Thanks scribs   


I'm also wondering why it matters to have the sequences ordered in that way. Irrespective of the order in which you add sequences to that original pair, won't the consensus sequence and the sequence logo be the same? If EQRK and EQRR are most similar in a group of 4 sequences...

EQRK       
EQRR             
AERR
QQQQ

vs

EQRK       
EQRR
QQQQ
AERR

Frequency of residues at each position is the same for each alignment ie consensus sequence is the same. Sequence logo should also be the same. hmmmm

appreciate any advice you can give me.

[scribble]

When you're trying to match your sequence with billions of sequences in a database, you want the sequences with the best match (lowest e value) to come up first. That's what the ordering is for.

[MelonBar]

From the flipped class:

Why can't mapping studies always pinpoint exact disease genes?

Any ideas?


Also.. what's the deal with HeLa cells? Can you just put genes in it and express it?

[mahler004]

From the flipped class:

Why can't mapping studies always pinpoint exact disease genes?

Any ideas?


Also.. what's the deal with HeLa cells? Can you just put genes in it and express it?

(IIRC) it's something along the line of multiple genes in one locus. GWAS studies may not have the resolution to pinpoint an exact gene, but may be able to narrow it down to say, six (which can then be sequenced).

HeLa cells are a pretty common immortal cell line. They can definitely be used for gene expression (using transfection procedures which aren't at all covered in BCMB30002). You can use them to test ideas about gene expression/investigate the sub cellular location of a gene using techniques covered in the course.

[MelonBar]

(IIRC) it's something along the line of multiple genes in one locus. GWAS studies may not have the resolution to pinpoint an exact gene, but may be able to narrow it down to say, six (which can then be sequenced).

HeLa cells are a pretty common immortal cell line. They can definitely be used for gene expression (using transfection procedures which aren't at all covered in BCMB30002). You can use them to test ideas about gene expression/investigate the sub cellular location of a gene using techniques covered in the course.

Thanks man appreciate it.