|
|
MOTU_define.pl
Frequently asked questions (and answers)
What is a MOTU?
Input files for MOTU_define.pl
The base difference cutoff
Why do different runs give different clustering?
How do I parse the output?
What has changed recently?
Input files for MOTU_define.pl
The input format for MOTU_define.pl is fasta formatted DNA sequence. Fasta format looks like:
>sequence name
AGCGGTGGCGTGGCGGTGGCGGTGGCCGGTG
AGCGGTGGCGTGGCGGTGGCGGTGGCCGGTG
AGCGGTGGCGTGGCGGTGGCGGTGGCCGGTG
All the sequences must be in separate fasta files. The programme copes with non-A/G/T/C base calls (all are coded "N").
The base difference cutoff
MOTU_define.pl uses a simple difference cutoff to place sequences into MOTU. The cutoff is set by the user, and can be varied in different runs to
explore the effects of different cutoff values on clustering. MOTU_define.pl does not explicitly use a tree to define MOTU, though of course the
use of BLAST implies that some phylogenetic relationship is expected.
Why do different runs give different clustering?
MOTU_define,pl uses single-linkage clustering: if a sequence is close enough (given the cutoff) to another, they are clustered. In some cases thie
order of analysis of sequences can change clustering. Consider three sequences, A, B and C. A differs from B by 2 changes and B from C by 2 changes.
C differs from A by 4 changes. If we use a cutoff of 2>2 differences, the order in which we add sequences will change clustering: A then C then
B and C then A then B will yield two clusters as A and C will not be linked. Any other clustering will yield one. The pattern of distribution of
different clustering can yield interesting information concerning "clouds" of closely related specimens that do not robustly form distinct
MOTU under specific parameters.
How do I parse the output
We have not implemented comparison routines in the software, but the output is easily analysed in, for example, Excel-type spreadsheets, or with
simple perl scripts. We will be placing some of these perl scripts on this site soon.
|
|
Website Highlight
The dog heartworm Dirofilaria immitis.
Filarial nematodes are tissue and gut parasites of a wide range of vertebrates, including humans. This species is a canine parasite and gets its common name of "heartworm" because the adults reside in the heart. It is closely related tospecies, such as O. volvulus, that cause human diseases, affecting over 120 million people. See NEMBASE4 for analyses of ESTs from this parasite and many other nematodes.
|
|
|
