This work is licensed under a Creative Commons Attribution 4.0 International License.
Prepped Aphaenogaster samples for shipping to BMGC for sequencing.
After correcting velvet assembler to specify that data format was fastq (not fasta), de novo assembly went quickly, ~ 1 hr run time!
Found online tool velvet advisor to help with k-mer selection. Based on my inputs of 12.5 million reads, paired-end, each read 60bp long, estimated genome size of 9 Mbp, recommended k=53. For 30 fold k-mer coverage, recommended k=51
Results for different k-mers. Assembly length is for contigs > 999 bp
| k-mer | num_contigs | 50% coverage quantile | assembly length contigs >999 (Mbp) |
|---|---|---|---|
| 21 | 13,853 | 88.2 | |
| 31 | 3,065 | 90 | |
| 41 | 1,695 | 71 | |
| 53 | 1,946 | 45 | 8.16e6 |
velvetOptimiser looks like a nice defensible way to go about selecting k-mer…but not implemented in Galaxy.
Another good reason to (finally!) migrate to command-line.
This work is licensed under a Creative Commons Attribution 4.0 International License.