Daniel:Notebook/GenomeMiner/2013-9-20

From ZhangLabWiki
Jump to navigation Jump to search

Mock HL155 (Started 9/9/2013)[edit]

Back to Calendar

Mimic Data, Bases 30 to 50[edit]

Last mimic data bases 30 to 50 actually put insertions/deletions from bases 1 to 29, not 30 to 50, so for good measure I'm repeating the code using the correct 30 to 50 range. This will make it directly comparable to the perfect base quality mimic data from 9/18.

Workflow[edit]

1. MockHL155_Master.m, Switch 5
2. scp v4s1mockseq_errormimic_30to50bp.fq djacobse@132.239.135.41:/media/LTS_15T/DEJ_LTS/SeqStore/130628_HL155/mockseq/errormimic_30to50/
3. ./hl155bash.sh
4. perl imp_count_mismatch.plx

Matlab Error Counts[edit]

627618 substitutions (0.42 pct) 
161680 insertions (0.11 pct) 
30992 deletions (0.02 pct) 

Alignment Results[edit]

2959000 reads; of these:
 2959000 (100.00%) were unpaired; of these:
   7858 (0.27%) aligned 0 times
   2945510 (99.54%) aligned exactly 1 time
   5632 (0.19%) aligned >1 times
99.73% overall alignment rate

Error Counting Results[edit]

Error Rate: 0.521%
Error Rate of Insertions: 0.090%
Error Rate of Deletions: 1.63e-04%
Error Rate of Substitutions: 0.431%

Full Reference Sequence[edit]

Just like yesterday, I'm using the real bases for deletions, instead of random bases at the end. This newest code should also be the fastest code yet, and boasts one more potentially important change. All other iterations (except the single error per read) would roll for errors in every base, including those that already had errors. The new code only gives potentially one error per base, meaning a base that has a substitution will not also be deleted (and other such errors).

Workflow[edit]

1. MockHL155_Master.m, Switch 7
2. scp v4s1mockseq_extseq_errormimic.fq djacobse@132.239.135.41:/media/LTS_15T/DEJ_LTS/SeqStore/130628_HL155/mockseq/extdeletions/
3. ./hl155bash.sh
4. perl imp_count_mismatch.plx

MATLAB Results[edit]

1493592 substitutions (0.52 pct reads) 
384892 insertions (0.13 pct reads) 
73602 deletions (0.03 pct reads)

Alignment Results[edit]

2959000 reads; of these:
 2959000 (100.00%) were unpaired; of these:
   688259 (23.26%) aligned 0 times
   2266933 (76.61%) aligned exactly 1 time
   3808 (0.13%) aligned >1 times
76.74% overall alignment rate

Error Counting Results[edit]

Error Rate: 4.301%
Error Rate of Insertions: 2.79%
Error Rate of Deletions: 0 %
Error Rate of Substitutions: 1.51%

These results seem really weird. I'm going to have to look over the Switch 7 code again.