loneknightpy / idba

124 stars 53 forks source link

IDBA-Tran and IDBA-UD generate segmentation faults and "invalid insert size" error. #7

Open jcornel3 opened 8 years ago

jcornel3 commented 8 years ago

Hello, the first time I ran IDBA-Tran I received several "invalid insert size" errors as shown here:

The command I ran: /home/jcornel3/tools/idba-1.1.1/bin/idba_tran -r /scratch/Xlav/RNA/data_for_build/normalizedk20.fasta -o ./ --num_threads 32 --max_isoforms 1

Log of the output: number of threads 32 reads 116728374 long reads 0 reads 116728128 long reads 0 read_length 90 kmer 20 kmers 1251934185 1281430428 merge bubble 753628 contigs: 87112541 n50: 34 max: 1340 mean: 33 total length: 2876019569 n80: 21 contigs: 13350365 n50: 91 max: 3603 mean: 51 total length: 691769806 n80: 23 contigs: 321705 n50: 417 max: 3603 mean: 429 total length: 138028498 n80: 337 aligned 16215662 reads confirmed bases: 116106490 correct reads: 1586511 bases: 185900 distance mean 149.455 sd 171.931 seed contigs 651896 local contigs 26700730 kmer 30 kmers 1352385351 1355369574 merge bubble 261359 contigs: 48240381 n50: 61 max: 5415 mean: 56 total length: 2721032711 n80: 33 contigs: 9744387 n50: 173 max: 8115 mean: 124 total length: 1209648933 n80: 87 contigs: 1384743 n50: 632 max: 14978 mean: 600 total length: 831992867 n80: 395 aligned 41531630 reads confirmed bases: 325970532 correct reads: 4992476 bases: 483422 distance mean 146.052 sd 444.039 invalid insert distance kmer 40 kmers 1314925458 1309923734 merge bubble 53515 contigs: 31990658 n50: 89 max: 5415 mean: 79 total length: 2528490281 n80: 48 contigs: 9324460 n50: 180 max: 8125 mean: 150 total length: 1407613381 n80: 92 contigs: 1665614 n50: 787 max: 14936 mean: 689 total length: 1148295112 n80: 430 aligned 49476710 reads confirmed bases: 390329809 correct reads: 5972961 bases: 183994 distance mean 124.625 sd 330.759 invalid insert distance kmer 50 kmers 1225819992 1218220098 merge bubble 10262 contigs: 21377438 n50: 112 max: 5415 mean: 101 total length: 2168392996 n80: 63 contigs: 7570936 n50: 206 max: 8135 mean: 180 total length: 1366986053 n80: 118 contigs: 1672231 n50: 851 max: 21594 mean: 720 total length: 1204534803 n80: 445 aligned 54056339 reads confirmed bases: 427486046 correct reads: 6452478 bases: 106997 distance mean 149.159 sd 415.923 invalid insert distance kmer 60 kmers 1064552217 1057789314 merge bubble 3040 contigs: 15286308 n50: 140 max: 5415 mean: 123 total length: 1887680552 n80: 76 contigs: 6281749 n50: 228 max: 7992 mean: 205 total length: 1289601622 n80: 138 contigs: 1522085 n50: 798 max: 24294 mean: 691 total length: 1052547585 n80: 431 aligned 56603526 reads confirmed bases: 447139164 correct reads: 6715696 bases: 58809 distance mean 160.791 sd 427.802 invalid insert distance

On subsequent attempts, I tried some different settings and now I just get segmentation faults:

Command: /home/jcornel3/tools/idba-1.1.1/bin/idba_tran -r /scratch/Xlav/RNA/data_for_build/normalizedk20.fasta -o ./ --num_threads 32 --pre_correction

Output: number of threads 32 reads 233456750 long reads 0 reads 224304660 long reads 0 read_length 91 kmer 40 Segmentation fault (core dumped)

Command: /home/jcornel3/tools/idba-1.1.1/bin/idba_tran -r /scratch/Xlav/RNA/data_for_build/normalizedk20.fasta -o ./ --num_threads 16

Output: umber of threads 16 reads 233456750 long reads 0 reads 224304660 long reads 0 read_length 91 kmer 20 Segmentation fault (core dumped)

Other people have used IDBA-UD and encountered a similar issue, here is a link to the google group thread of the issue: https://groups.google.com/forum/#!topic/hku-idba/Y-7dLOcMXvU

The other thread noted that the reads need to be formatted correctly. I'm fairly certain mine are formatted correctly since I used this dataset with SOAPdenovo-Trans and it worked well.