Why do human genome studies use bwa 0.5.9?

[wfefw]

Many papers in human genetics including published in this year have used bwa 0.5.9 for HiSeq reads mapping instead of upper versions.

Why do they use the old one?

[Bukowski]

Because the work was probably done some years ago and it's only just been published. There's quite a lag between sequencing and publication - possibly the lifetime of a 3 year grant.

And the performance on 100 base reads probably isn't that different to later versions anyway.