best way to index tab-delimited text file

[Joseph White]

What is the best way to index a tab-delimited text file containing chromosome, position and variant data? My files are huge and too big to maintain in memory, so indexing seems the only viable option.

jwhite

[GenoMax]

tabix from Samtools.

[Joseph White]

tabix from Samtools.

The file is not in BED, GFF, or SAM format.

[GenoMax]

Tabix can accept these formats: -p gff|bed|sam|vcf

[Joseph White]

Tabix can accept these formats: -p gff|bed|sam|vcf

Oh, that's right. All I have to do is add a third column of '.' to make it VCF. Thanks.

[finswimmer]

Hello,

Tabix can accept these formats: -p gff|bed|sam|vcf

these are just presets. One can define in which column the chromosome (aka sequence name), begin and end position are located.

So if you have the chromosome name in the first column, the position (begin == end) in the second column you can index like this:

tabix -s1 -b2 -e2 my_file.gz

This way, tabix provide a way to index each tab delimited file, which have sorted positional data. Also one can define whether the position is 0-base or 1-based give the parameter "-0" if it's 0-base.

fin swimmer

[sam657]

These methods are still working for you?

[finswimmer]

These methods are still working for you?

Yes, why they shouldn't? It's a documented feature. The -p parameter is just a shorthand for this. So in the case above, it would also work to use "-p vcf" as the chromosome name is column 1 and the position in column two like it is in a vcf. About the other columns tabix don't care. It doesn't check whether it is a valid vcf file.

fin swimmer