filter.seqs

Commands in mothur
1

I think a loose a lot of lenght in the filter.seqs command.
This is the result of the summary.seqs before screen.seqs
start end Nbases ambig polymer numseqs
minimum 1044 1046 1 0 1 1
2,5% title 6237 16427 249 0 4 356
25% title 6428 22612 403 0 5 3552
median 6428 25298 406 0 5 7103
75% title 6430 25298 423 0 5 10654
97,5% title 7193 25298 430 0 6 13850
maximum 43116 43116 497 0 7 14205
mean 6633,32 24105,4 401,895 0 5,03464
unique seqs 11761
total seqs 14205

then I run screen.seqs with the parameters optimize=start-minlength, criteria=85 and I get:

start end Nbases ambig polymer numseqs
minimum 5707 22091 391 0 4 1
2,5% title 6232 22580 395 0 4 258
25% title 6428 25298 404 0 5 2575
median 6428 25298 407 0 5 5149
75% title 6428 25298 423 0 5 7723
97,5% title 7192 25298 431 0 6 10040
maximum 7192 26918 497 0 8 10297
mean 61641,64 413,055
unique seqs 8440
total seqs 10297

will here ok. Then I run screen.seqs(fasta=current, vertical=T, trump=.) and then I get
start end Nbases ambig polymer numseqs
minimum 1 780 304 0 3 1
2,5% title 1 784 312 0 4 258
25% title 24 784 323 0 4 2575
median 24 784 324 0 5 5149
75% title 24 784 342 0 5 7723
97,5% title 25 784 359 0 6 10040
maximum 27 784 371 0 8 10297
mean 22,0251 783,993 329,348
unique seqs 8440
total seqs 10297

the mean Nbases has reduced from 413 to 329 Pb. is this normal? What’s happening?

Thanks in advance.

2

why not try…

screen.seqs(fasta=, group=, name=, optimize=start, criteria=85, end=25298)

because sequences vary in their number of insertions and deletions within the same alignment region, sequence length and alignment length aren’t necessarily the same thing.

Pat

3

Thank you.
I will try that, but I really don’t like much because I was sequencing from the forward primers, so sequences should start at the same pont, not end at the same point.
Anyway, there is something I don’t understand. Theorically, when I run filter.seqs(fasta=, vertical=T, trump=.) I am eliminating just columns with a “-” in all the sequences and the columns that present a "."in some sequence.By this I understand that I am not reducing the number of Nbases, just the length of the alignment, but it reduces in more than 100pb the number of bases.

By the way, when is a “.” inserted in a sequence, I read that it means missing data, but I don’t really understand the meaning.

Thank you.

4

Hi Anna,

You can choose your start position instead of your end position, e.g. ‘start=19854’ instead of ‘end=25298’.

The ‘.’ means that it it not known what nucleotide/gap follows or precedes the sequence. By removing all columns with a period (which should only be found at the ends of sequences), you will inevitably chop off portions from the sequences with longer lengths.

Cheers,
Stephen