Looking for DNase-seq of MCF10A cell line

Question

[bhchiang]

Where can I find data from a DNase-seq assay of the non-cancerous MCF10A breast cell line aligned to the hg19 genome?

I've found the following:

• MCF10A-ER-Src (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSM2676283)

• MCF10A aligned to hg38 with only one biological replicate (https://www.encodeproject.org/experiments/ENCSR789VGQ/)

  • Is there any way I can align this to hg19?

• A study using MCF10A DNase-seq but is not posted on GEO

  • https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE84579
  • https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE84579

• FAIRE-seq of MCF10A

  • https://www.ncbi.nlm.nih.gov/sra/?term=mcf10a+faire-seq

Any help would be appreciated, thanks so much!

Answer 1

[Kevin Blighe]

•MCF10A aligned to hg38 with only one biological replicate (https://www.encodeproject.org/experiments/ENCSR789VGQ/)

◦Is there any way I can align this to hg19?

Yes, download the raw FASTQ files and then re-align them to hg19.fasta using Bowtie or some other aligner.

Your third listed study ( https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE84579 ) does not have MCF10A, from what I can see.

Your first study ( https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSM2676283 ) was treated with Tam.

Kevin

[bhchiang]

Would it be possible to simply perform CrossOver on the hg38 MCF10A DNase-seq coordinates instead of realigning them? Or will realigning them be more accurate? Thanks!

[Kevin Blighe]

I would definitely re-align them, even with the added computational expense in doing this. Crossing-over from one build to another can be tricky because many regions will exist in one build but not another. It is a potential mine-field...

[bhchiang]

Got it. I'll give it a shot, thanks. If possible, could you link me to any tutorials you've useful for aligning DNase-seq files?

[Kevin Blighe]

I feel encouraged that the HOMER authors have specifically addressed DNase-seq, even warning of the differences between FAIRE- and DNase-seq, which look at the same thing but via different protocols: DNase-Seq Analysis Tutorial

HOMER refers to these as the original 'Crawford' method (DNase-seq) and the 'size selection' method (FAIRE-seq). I looked at the MCF10A protocol (https://www.encodeproject.org/documents/926174f5-d14c-4e77-bc52-5517b56daac0/@@download/attachment/Culturedcells_SOP_nuclei_DNase_crosslink_RNA_V1.pdf) and I am fairly confident that they used the DNase-seq 'Crawford' method.