Novoalign options

[Gen2007]

Hello Group,

I've been trying to find the optimal parameters for aligned bisulfite treated 76bp PE reads to a small reference that includes repeats. I'm confused with the option to set the fragment length and standard deviation. My question is: what exactly does the fragment length refer to? Is it the distance between mapped mates or does it include the reads.

The library I am using isolated adapter fragments between 250-300bp (minus adapter = 131-181). Given 76 bp reads, some mates could overlap. If the fragment length refers to distance between aligned reads than this value could be negative. Anyone want to clarify?
Thanks a lot!

[sparks]

HI, Just found your post. The fragment length mean and standard deviation are outer coordinates, it doesn't matter if you set them a bit higher than the real values.
Colin

[Gen2007]

So by outer coordinates you mean the most distal coordinates of an adapter-ligated molecule, i.e. average length of your library that you would see after PCR amplification on a gel?

For the library I mentioned, I used 150bp as the mean fragment length since I assumed it was the distance between reads, and the alignment worked pretty well. Is it more computationally strenuous to have a larger fragment length setting?

[sparks]

It's the length of the DNA fragment as mapped by the aligner and hence doesn't include any adapters. It's not the length of the gap between the two read alignments.
If your gel includes PCR adapters then you need to adjust for this.

It doesn't really affect computation as Novoalign adjusts to the fragment lengths it sees, but setting it too short may mean reads don't get paired properly.
It's not a major issue as range for pairs is from 0 to mean + 6 standard deviations so it's usually enough.
You can also run Novoalign on a few K reads and check the reported fragment length distribution. Use the -# option to limit the number of reads processed. e.g. -# 2K will map 2000 reads and stop.

[Gen2007]

Great, that makes a lot of sense. I actually meant "reads" instead of "adapters." Of course in Illumina sequencing the read begins before the adapter since the sequencing primer anneals to the adapter... I'm glad you caught that. I appreciate all the help. Thanks a lot!