UCSC reference assemblies (eg. hg38) are, with some minor differences, almost the same as GRC references. If they have nothing new to offer, why bother making and releasing reference genomes? I'm not criticizing UCSC, I'm just curious about the institutional politics or bureaucratic decision-making behind US taxpayers funding two nearly identical projects?
Wouldn't it be in the scientific community's interest to move to one common standard? NCBI vs EBI I can understand because of divergent national interests but since UCSC and GRC are both US-based and publicly funded, why isn't there central coordination to move to one or the other standard?
•
written
by
BioinformaticsLad
2 answers
UCSC doesn't produce the reference genomes, but rather collects them with a browser and adds annotation files (which also are not necessarily produced by them). Annotation is at least as important as the assembly itself.
From a quick look at Wikipedia, GRC is international:
The UCSC Genome Browser is an on-line, and downloadable, genome browser hosted by the University of California, Santa Cruz (UCSC).
The Genome Reference Consortium (GRC) is an international collective of academic and research institutes with expertise in genome mapping, sequencing, and informatics, formed to improve the representation of reference genomes.
Log in to answer this question.
You're obviously not from the US, we don't centrally coordinate much of anything in any coherent way. Note also that GRC handles only mouse and human and carried on work by UCSC...
Correct, I'm not from the US. If a top-down approach is impossible in the US system, can't researchers from a bottom-up manner collectively move to one standard? Or is even suggesting using common standards too much to ask for in bioinformatics? This isn't so much an issue of "which one is better" but rather "just pick one and stick to it".
That's too much to ask for from our species in general.
I would say we're doing fairly well with common standards in bioinformatics, as most genomes are in fasta, reads in fastq, alignments in sam/bam/cram and variants in vcf, etc. Having multiple sources of reference genomes is maybe inconvenient, but not terrible.