Trinity and SOAP de-novo used the default k-mer setting [ 40, 42], and trans-ABySS was run using k-mer values of 45, 46, to 89 [ 41].
We also conducted assemblies using both the merged and unmerged reads with Velvet version 1.1.02 [ 39] and k-mer values of 71, 81, and 91.
In particular, we used k-mer values of 51, 61, 71, 81, and 91 and varied the coverage (c) and erode (e) parameters from 2 to 1,000.
First, in order to build highly represented transcripts, 10 million random reads were assembled using ABySS with k-mer values of 35, 45, 55, 65 and 75.
We used K-mer values of 29 for the six samples with single reads and 31 for the remaining samples with paired-end reads.
K-mer values of 51, 61, 71, and 81 were used, followed by the removal of redundant contigs with CD-HIT-EST (Li and Godzik 2006).
Oases was run by using k-mer values of 21, 23, 25, 27 and 29, and Soaptrans was run by the same strategy using k-mer values including 37, 41, 45, 49 and 53, while IDBA-tran was run by using k-mer values of 41, 43, 45, 47 and 49.
All reads mapping to these selected scaffolds/contigs using SOAP 2.21 [ 62] were extracted, along with their mates, and then assembled using CLC with k-mer values of 23, 35, 51, or 63.
Genome wide scans for repetitive elements in the mtDNAs from S288C, NCYC3594 and YJM789 were performed utilizing Jellyfish [ 79] with k-mer values of 15, 30, 50, and 100.
Briefly, reads remaining after quality control were used to construct de Bruijn-graphs with k-mer values of 21, 31, 41, 51 and 61 using Oases v.0.2.09 [ 9].
The assemblies were performed with k-mer values of 25, 29, 33, 37, 41, 45 and 49, and the assembly with k-mer value 41 was chosen based on assembly statistics and the presence of known sequences.
