I want to compare the pre-mRNA alternaive splicing events between RNA-
seq datasets. Do I need to allow indel search when I run Tophat? What
is the indel search for? I could not find detail information about
"indel search" through the documentation of Tophat.
In simple terms, "No" produces a strict alignment and "Yes" produces a
more permissive alignment.
The option 'Allow indel search:' is a way of allowing for variability
your data (presumably biologically valid) to not be interpreted as
mismatches or gaps. Mismatches/gaps in an alignment lower the overall
score and can lead to alignment failures. The default for this
is "Yes" with value of 3 for insert/deletion length in Galaxy
for simple nucleotide polymorphism variability up to a single codon,
position, in either the query or target). All values can be modified.
If this interferes with your data mapping accurately, then it could be
disabled by setting the parameter to "No". A test comparing the two
alternatives on a sample would be a good way to see how this single
change affects your particular sample. Good questions to ask: What
do not map when the stricter alignment rules are applied? Do any reads
map with a change in specificity? Do you agree with the results?
Hopefully this helps!