Comparison of different number of threads to run pindel
note
- Download pindel from https://github.com/genome/pindel at 2016/05/25
- Test data: pindel-master/demo
- Pindel version 0.2.5b8, 20151210.
- Intel® Xeon(R) CPU E5-1620 v2 @ 3.70GHz × 8
- Mermory 16G
- Ubuntu 15.04
1 thread
1 | time ../pindel -i simulated_config.txt -f simulated_reference.fa -o bamtest -c ALL # 14.95s user 0.10s system 99% cpu 15.066 total |
4 threads
1 | time ../pindel -i simulated_config.txt -f simulated_reference.fa -o bamtest -c ALL -T 4 # 23.92s user 0.11s system 285% cpu 8.429 total |
result data download here
Compare
md5
Do md5sum *
in result_1
and result_2
md5 in result1 | md5 in result2 | result file |
---|---|---|
d41d8cd98f…… | d41d8cd98f00…… | bamtest_BP |
d41d8cd98f…… | d41d8cd98f00…… | bamtest_CloseEndMapped |
8601dc53a8…… | 8601dc53a8ec…… | bamtest_D |
d41d8cd98f…… | d41d8cd98f00…… | bamtest_INT_final |
a92f5b2110…… | 1fd2fe1eb99f…… | bamtest_INV |
d41d8cd98f…… | d41d8cd98f00…… | bamtest_LI |
653f906ec7…… | 421062310b1b…… | bamtest_RP |
e52cdc04a9…… | 3aa3fa3ddd19…… | bamtest_SI |
0017be684e…… | 0017be684e37…… | bamtest_TD |
diff in string
Using Beyond Compare
bamtest_INV
Just some different order
1 | ATTTGGATCTTGGGCACAGGGACCTCAACCGACAAACCTAGCATGTCATAAGTTCGTGGCACTATGAAGCTTTCCCTATAAACCAGTCCACTCAAGTGAT + 49528 0 SAMPLE1 @49848/1 |
1 | ATTTGGATCTTGGGCACAGGGACCTCAACCGACAAACCTAGCATGTCATAAGTTCGTGGCACTATGAAGCTTTCCCTATAAACCAGTCCACTCAAGTGAT + 49528 37 SAMPLE1 @49848/2 |
bamtest_SI
A little insignificant difference (SUM_MS and order)
1 | 8 I 44 NT 44 "CACGCCCATGTGCTAACTTGGTGAGCATCACTAAAACTACAACC" ChrID 3 BP 50198 50199 BP_range 50198 50201 Supports 2 2 + 0 0 - 2 2 S1 3 SUM_MS 74 3 NumSupSamples 1 1 SAMPLE1 0 0 0 0 2 2 SAMPLE2 4 4 0 0 0 0 SAMPLE3 4 4 0 0 0 0 |
1 | 8 I 44 NT 44 "CACGCCCATGTGCTAACTTGGTGAGCATCACTAAAACTACAACC" ChrID 3 BP 50198 50199 BP_range 50198 50201 Supports 2 2 + 0 0 - 2 2 S1 3 SUM_MS 37 3 NumSupSamples 1 1 SAMPLE1 0 0 0 0 2 2 SAMPLE2 4 4 0 0 0 0 SAMPLE3 4 4 0 0 0 0 |
bamtest_RP
sort it by coordinate: $ sort bamtest_RP -k2 -n > bamtest_RP_sort
some different in content and deviation in coordinate
conclusion
In my opinion, multiply threads may have little influence in SVs but RP
result data download here