Transcriptome assembly with digital normalization and read merging

Completed velvet-oases assembly on digitally-normalized reads.

About 1hr 15min time for assembly…FAR faster than velvet on all reads with much lower memory usage.

Summary with python /opt/software/khmer/sandbox/assemstats2.py 100 transcripts.fa

diginorm only: A22-diginorm-oases/A22-oases-21/transcripts.fa

assembly stat	result
Total Contigs	126812
Total Trimmed Contigs	126781
Total Length	109476821
Min contig size	100
Median contig size	365
Mean contig size	863
Max contig size	14314
N50 Contig	16370
N50 Length	1933
N90 Contig	66842
N90 Length	333

diginorm then FLASH: A22-diginorm-flash-oases/A22-oases-21/transcripts.fa

assembly stat	result
Total Contigs	111434
Total Trimmed Contigs	111413
Total Length	111343478
Min contig size	100
Median contig size	447
Mean contig size	999
Max contig size	20427
N50 Contig	15236
N50 Length	2163
N90 Contig	58158
N90 Length	410

Compare to previous results running FLASH first, then diginorm; both with minimum contig length of 100 bp.

FLASH before diginorm: A22-oases/A22-oases-21/transcripts.fa

assembly stat	result
Total Contigs	90612
Total Trimmed Contigs	90612
Total Length	86485229
Min contig size	119
Median contig size	586
Mean contig size	954
Max contig size	14006
N50 Contig	16436
N50 Length	1506
N90 Contig	60314
N90 Length	396

40% more contigs and 26% greater total length in assembly with diginorm alone.

Still waiting for full assemly using velvet with 200GB RAM and 48hr walltime to complete for comparison, but at least in the number of contigs, the results are consistent with those from VGN with full Trinity assembly.