sequence quality values

darked at darked at
Fri Oct 27 11:11:41 EST 2000

Dear All,

I run sequencing core samples on ABI377XL sequencer (10h runs, 48cm long
plates, BigDye kit). I do basecalling on unix using phred version
As it may be expected, the sequence quality varies very much from user
to user. I need to have a numeric measure of sequence quality other than
visual inspection of all traces. So far I've looked at two values,
signal strenght (SIGN extracted from SCF files) and AQ value (AVerage
Quality for bases 100..200 extracted from exp files created by
phred/pregap4). Both values seems to be unsatisfactory to make
discrimination between good and bad quality samples straightforward.

I wonder if somebody has been able to come up with better measurements
of sequence quality and can share them with me. Many thanks. :-)


Darek Kedra darek.kedra at

My example values:
ND_000815c01.exp:AQ   74.520000
ND_000815c01.scf SIGN=A=59,C=35,G=54,T=28
visual_inspection: good quality

ND_000815c02.exp:AQ   64.370000
ND_000815c02.scf SIGN=A=26,C=24,G=23,T=20
visual_inspection: close to random noise

ND_000815c03.exp:AQ   81.780000
ND_000815c03.scf SIGN=A=69,C=31,G=46,T=30
visual_inspection: good quality

ND_000815c04.exp:AQ   83.800000
ND_000815c04.scf SIGN=A=87,C=37,G=56,T=34
visual_inspection: good quality

Sent via
Before you buy.


More information about the Autoseq mailing list