hbctraining / intro-to-chipseq Goto Github PK

https://bioconductor.org/packages/release/bioc/vignettes/regioneR/inst/doc/regioneR.pdf

We don't currently have an extensive amount of metadata about the data as we do with RNA-seq. We can find out this information and add it in

I think we can either create new or find better figures to represent that

• Lecture on alignment like we do in RNA-seq? Yes. - A shortened version (i.e. remove the suffix array stuff).
• Add in BWA and ask Rory about including some benchmark results.

this might be good since we have removed trimming

We've been using csaw for histone modification studies as an alternative to peak-calling or focusing on specific features like TSS regions. It might be useful as a lesson.

rather than have them install ChIPQC, spp, catools, and any other R packages this might be easier

Pertaining to ChIPseq lesson 01

• check everything for flow
• check language

change this as we did with RNA-seq, which means removing the "best practices" section in the QC lesson

• Talk in more detail about the SAM file
• What information is stored in it etc. Take from RNA-seq

move the lessons from Intro to RNA-seq (O2)

• (Host says orchestra)

From Rory:
"greylist regions are regions where the input exceeds a threshold, where peak-callers sometimes call spurious peaks. threshold is calculated by calculating depth over the input, sampling it repeatedly and estimating negative binomial parameters and then taking the threshold as the .99 quantile of the NB"

https://github.com/roryk/chipseq-greylist i just copied what the chipseqgreylist R package does

Check with Rory. We want to match what is happening in bcbio

The current script incorrectly mentions that duplicates are removed during one of the sambamba steps, this needs to be corrected.

• create a new directory on O2 where raw_fastq contains the chipseq files
• change mentions of rna-seq
• specifically modify the data organization markdown (merge with Intro to Unix)

remove 1-3

currently in my home directory

Even the move to ChiP

remove trimming since we are soft-clipping with bowtie2?

• Samtools is being used in QC lesson, needs more detail to introduce it.
• Add a note on similarities between samtools and sambamba. Only using sambamba where necessary.

and ask Rory about including some benchmark result

Need to check this since Kayleigh said it didn't. Also if it does, does it still contain a cross-correlation plot?

When talking about local alignment discuss that this is the soft-clipping described in the QC section.

This is an HTML tag taht would be useful for "try it on your own" sections, where the code is hidden initially but clicking on it will make it available

this is how you would do it for your own dataset anyway. Or hopefully get it running on O2 (fix X11 problem)