Created the following tickets to move forward with muday plans:

IGBF-3436 Create DESeq Genotype and Time Designs for Muday-144
IGBF-3437 Create EdgeR Genotype and Time Designs for Muday-144
IGBF-3438 Compare Muday-144 Deseq and EdgeR results and perform a sanity check
IGBF-3439 Create top 25-100 DE genes output files for Muday-144
IGBF-3440 Use top 25 DE genes from Muday-144 Analysis and visualize with original scaled counts file to make lines graphs and/or heat maps

Moving this ticket to done!

Next Step for tickets:
1. Create the designs for new DESeq and EdgeR analysis'

• review design creation: https://youtu.be/X6p3E-QTcUc?t=646
• ensure the designs answer Muday Labs questions about genotype and time
• make sure to use the same designs for DESeq and EdgeR
• run the analysis for both statistical methods twice for genotype and time

2. Compare Deseq and EdgeR results and perform a sanity check
3. Create top 25-100 DE genes output files
4. Use top 25 DE genes and visualize with original scaled counts file to make lines graphs and/or heat maps

Question: Do we want a separate ticket for EdgeR and Deseq for each step?

I think we should use edgeR for this because the documentation is a lot better.

See: https://bioconductor.org/packages/release/bioc/vignettes/edgeR/inst/doc/edgeRUsersGuide.pdf.

Next step: Create tickets for genotype + temperature and time + temperature analysis.

Notes from meeting: Also spoke about later maybe using heat maps to see gene expression over time for possibly top 20 significant genes. The Muday lab was going to look into Blast to-go and see if they want to use it for annotations.
Questions that were asked: Group was curious why there were less down-regulated genes than up-regulated on the temperature designed volcano plots. Curious also about why there were more DE genes for ARE compared to the other genotypes.

New ticket for SL4 names: IGBF-3407

Change request for [~molly]: the work you did creating the new table mapping SL4 to SL5 names is out of scope for the task of this ticket, which is to plan a new analysis. It would be better to create a new ticket for just that one task and track it there, instead of here.

Another change request: Please look at the other tickets associated with this epic. Some of them may be related. There may already have been some work done on this.

Gloria's request from her email:

Also, we have a small problem with the excel files on the DE genes within genotype and between treatments, that I think you can help us with easily.
We are trying to run an enrichment analysis and the software does not recognize the newest V5 genome. You gave us the S4 IDs, but they are part of the description, rather than a separate column. Is there a way to add this to the spreadsheet as a column? That way we can cut and paste a group without manually extracting each gene ID from that text heavy gene description list.
Thanks
Gloria

Meeting Prep notes:

DESeq performs an analysis based on the design you choose for it. For example, if I choose a time + temperature design you have chosen interactions to specifically analyze/compare time + temperature DE genes for each experiment/sample. If I just chose temperature for my design then the analysis will focus just on comparing 28C and 34C DE genes. So we can choose many different designs for DESeq and have single or multi factor designs.

Time Design =

dds <- DESeqDataSetFromMatrix(countData = round(cts),
                                colData = coldata,
                                design = ~ time + temperature) 

Genotype Design =

dds <- DESeqDataSetFromMatrix(countData = round(cts),
                                colData = coldata,
                                design = ~ genotype + temperature) 

Design video tutorial: https://youtu.be/X6p3E-QTcUc?t=645

Notes about previous Designs:

• PCA plots: time + temperature Design
• Volcano Plots: temperature Design

Questions:

• How would you like to visualize these interactions? Heatmaps, volcano plot, PCA plot, MA plot, plot top 20 of significant genes, or plot expression of a single gene.

Code Resources:

• https://hbctraining.github.io/DGE_workshop/lessons/06_DGE_visualizing_results.html
• http://bioconductor.org/packages/devel/bioc/vignettes/DESeq2/inst/doc/DESeq2.html

I have sent an email out to Gloria to arrange a meeting to discuss what products need to producing for this task.
We can then break out out into more tasks once we have buy in from Gloria.

Rob