Clarification sought on coordinate schemas and positions about libbigwig

[dpryan79]

The first base is at position 0 in a "zero-based half open" coordinate system. This is the same as for BED files and BAM files, for what it's worth. For example, the following specifies the first base in bed format:

chr1 0 1

I didn't come up with the format, this is just for consistency with it.

All bigWigs are always 0-based, the coordinates are never different. Fixed and variable-step files are also 0-based half open. This is different from the wiggle format, which always uses 1-based coordinates.

[andrewyatz]

Don't worry I'm not placing the sole responsibility for the coordinate system at your door. Far from it. I'm part of the Ensembl project and am fully aware of the UCSC coordinate scheme however there are times it catches me out.

As to your last point I believe I got my info from this blog post by the UCSC genome browser team. http://genome.ucsc.edu/blog/the-ucsc-genome-browser-coordinate-counting-systems/. Specifically the last bit concerning wiggle files. They way they had structured this it looked to me that fixed and variable step are stored in a BigWig as 1-start fully closed.

However if they're not then I can stop worrying and assume everything is 0-based, half-open that makes everything a lot easier to understand.

Thanks

[dpryan79]

I think they're just inconsistent in the wiggle format...which is crazy. To my knowledge bigWigs are always consistent here, or at least any produced by this library are.

[andrewyatz]

That's fantastic to hear!