bfast index

[kalinka23]

Hi everybody,

i am new here at this forum and hereby i would like to greet everybody.

I would like to try the bfast algorithm to map Solid reads (colourspace) to human genome. Now i try to index the genome (hg19) but i am really confused how to do that.

Did i understand it right that, if i want to use the recommended masks (given in the bfast-book, section 7.1.2 for ABI Solid reads with at least 50bp in length) that i have to index the genome 10 times using the following commands:

bfast index -f hg19.fa -m 1111111111111111111111 -w 14 -i 1
bfast index -f hg19.fa -m 111110100111110011111111111 -w 14 -i 2
bfast index -f hg19.fa -m 10111111011001100011111000111111 -w 14 -i 3
.....

bfast index -f hg19.fa -m 1110110001011010011100101111101111 -w 14 -i 9
bfast index -f hg19.fa -m 111111001000110001011100110001100011111 -w 14 -i 10

???

And another question is:
for the mapping the reads in colorspace are converted to fastq files and than used for the mapping (using the command: solid2fastq -n ... -o reads *.csfasta *.qual). Doesn't have the conversion from colorspace reads to fastq reads an adverse affect (in terms of e.g. loss of information ...)?

Thank you very much!

Best regards,
Karin

[nilshomer]

1. Yes, you must create ten indices.

2. No loss of information. The original data is present in the mapped SAM file in the CS/CZ tags.