I am working on a NGS data analysis project. I am into the computational domain of the problem and am completely new to the NGS domain. I need to decide whether to use input in ADAM format or in BAM format. I also need the data available chromosome-wise for further processing, for which I find BAM files better comparatively. I wanted to know the popularity and the future prospects of both the formats.
There is faintingly little usage of ADAM in bioinformatics, do not spend time adding support for it unless you want to target the tiny sliver of the community that has access to and can make full use of large clusters using apache spark.
Supporting BAM (and ideally CRAM too) will allow your application to be usable for the next 5 years at least.