Run AVA/AVA-CLI without needing to specify MIDs

[trasver]

Hi all

Is it possible to run AVA (specifically command line AVA) without needing to specify an MID when creating the project. We have sff files that correspond to a single sample and have no barcoding information in them yet still want to use AVA to map these reads to a reference sequence and identify variants.

Thanks for any suggestions

[vamosia]

Yes its possible, you can ignore the MID section. But you'll need to specify your primers (forward and reverse) as well as what your reference sequences are.

[Himalaya]

Yes its possible, you can ignore the MID section. But you'll need to specify your primers (forward and reverse) as well as what your reference sequences are.

Hi Vamosia
I tried ignoring the MID section and specified the primers and reference also. But still not working. Are you saying about sequencing data in sff containing no any MID sequence? We have a sff file containing all MIDs sequences and then the sequence reads are extracted with individual MID. Each MID extracted sff file is then tried to align with reference sequence providing all primer information (forward reverse with start stop position) but we still get error " No Sample/Reference to align". What's wrong? Any suggestion please.....

[maven]

If your original sff file has MIDs, why are you splitting them ahead of time? You should leave them alone and demultiplex in AVA. Besides, when you split using sffinfo, you get fasta files not sff files, so assuming you are splitting with sffinfo then you're either wrong in your description of what you're doing (i.e. you're trying to use fasta files with AVA), or you're incomplete (i.e. after splitting the sff file, you used the resulting information in the fasta files to generate a matching set of sff files). Anyway sff files include data for every flow, including key and MID, regardless of whether they have been subsetted from a precursor file. The only difference is where the 5' trim point is set. But AVA doesn't respect the 5' trim point in sff files. So if you are using AVA, by definition you have to use sff files, and therefore you also have to deal with the MID regardless of if it's split or not. So why split.

[Himalaya]

If your original sff file has MIDs, why are you splitting them ahead of time? You should leave them alone and demultiplex in AVA. Besides, when you split using sffinfo, you get fasta files not sff files, so assuming you are splitting with sffinfo then you're either wrong in your description of what you're doing (i.e. you're trying to use fasta files with AVA), or you're incomplete (i.e. after splitting the sff file, you used the resulting information in the fasta files to generate a matching set of sff files). Anyway sff files include data for every flow, including key and MID, regardless of whether they have been subsetted from a precursor file. The only difference is where the 5' trim point is set. But AVA doesn't respect the 5' trim point in sff files. So if you are using AVA, by definition you have to use sff files, and therefore you also have to deal with the MID regardless of if it's split or not. So why split.

Hi Maven ...
Firstly, the precursor sff file is split using sfffile tool, i think, so the files i got are sff file, not fasta file, after splitting using individual MID. Also the the trim point is set to the base after the MID sequence after splitting. So I am still submitting sff file to AVA after extracting reading with MID. It still fails giving same error....But if I provide the MID, no error. The only difference is slightly different number of reads aligned to reference...I mean to say, if i run AVA precursor sff file..for MID1 the number of reads aligned is lets say 1000 then the number of reads aligned for MID1 sff file after splitting will be a bit less than 1000. However this is the point with MID information supplied to AVA. But I want to run AVA for sff files after splitting...without MID given. Each MID is correlated to a single sample, so want to run for single sample at a time....I am curious how vamosia ( his post up in this thread ) run skipping the MID section.....Thanks Maven...Waiting for Vamosia if he puts some light on this....

[maven]

Himalaya, you are right - totally on me for providing misinformation. However as you've discovered, you will still have to configure the MID in AVA and use a multiplexer because the MID is still in the read. BTW the OP said the sample had "no barcoding information", which may be why Vamosia said you can ignore the MID configuration. Anyway, good luck and sorry for steering you wrong with my misstatement about sfffile vs sffinfo.