GWIPS-viz Forums

Full Version: Workflow differences between hg19 and hg38
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
I have a question about the procedures followed for mapping data to hg19 and hg38. Was the workflow same for both? If I understand correctly, the method described in Michel et al., 2014 was used for hg19 data. When new data were mapped to hg38, was the same workflow employed, including the use of RUM to map intron-spanning reads? The workflow published in RiboGalaxy ( does not have the RUM step. Was this done outside Galaxy?

For example, if we look at the example of a gene RPS23, hg19 data clearly supports translation initiation in exon 1 which encodes only the methionine ( However, looking at the hg38 data (, the initiating ribosomes' profile nearly mirrors that of the elongating ribosomes' profile. I imagine this would occur if some reads aligning to exons 1 and 2 were dropped when mapping to hg38 and such reads would be enriched in the samples treated with a drug to arrest ribosomes in the initiator site.
Hi there,

The short read alignment tool, RUM, was initially used to map reads in GWIPS-viz. However, when the developers informed us that RUM was no longer being developed, we switched to bowtie, which RUM used. This is why the more recent tracks have been mapped with bowtie and also why RUM is not included in RiboGalaxy.

Regarding the initiating data for RPS23, what you suggest is indeed a possibility. When we map the older datasets to the hg38 genome, we will be able to check this and let you know. However, it may also originate from differences in datasets - compare for example some of the individual tracks for hg19 ( and hg38 (

Reference URL's