alignement score is dependent on matches and mismatches
alignment score of 0 = 100% perfect match
1 mismatch = -6 score
1 ambinguous character = 1
if there’s a gap in the reference seq. or the read seq: 5 + 3*gap length
a 2bp gap in the read seq. = 2 insertions (relative to the read seq.)
a 2bp gap in the ref. seq. = 2 deletions (relative to the read seq.)
default in normal alignment mode is L, 0, -0.2 (where L = linear, the first number = the intercept, and the last number = the slope, and x = read length). The default specifies for reads that are 100bp the alignment score must be >= -20.
in –local alignment mode, G is used and G = natural log (i.e. f(x) = 1.0 + 5.4ln(x), where x is the length of the read). The min score default in local mode is 26 for 100 bp read.
my old alignment settings for the CrY2H-seq pipeline were:
–local mode
mismatch <= 2
indels <= 2
30bp match length
no softclipping in the beginning of a read because that is where the quality should be the best
Compare different alignment score thresholds (i.e.-0.6, -1.2, etc.)
look at % aligned
run dedup.
Caught up with Brent (brentv@uw.edu)
set up meeting time with Steven and Brent (Monday Oct.1 @ 1:30)
Pt. Whitney can bring in 80 total animals
possible conditions could include
an acute low treatment (i.e. pH shock for 1 month)
different collection times to expose different stages of gonad dev.
need to do a power analysis to see # animals needed and # of possible treatments