notes_for_next_course.txt

notes for next course

 O Slides trop chargee en general -> streamline for presentation 
    O 2 documents : slides et notes de slides
 O Trop de contenu pour certains, pas assez pour d'autres ?

 O Disconnect between theory and practice : 
    O focus more on results interpretation.
    O + stretch goals
 
 X cheatsheet specific to the course

 O be clearer about where the script should be launched with
    - imperfect solution in webpage
    - keep good solution on server (chemins absolu ... ) OR special tab in webpage

 O use tests for pre-requisites
 O keyboard-litteracy pre-requisite
 

 in R:
  O more complex model 
  O more matrices, try to go for group projects if there is time?



Course website : https://sib-swiss.github.io/RNAseq-introduction-training

general TODO:
   O clean slides : less text, clearer messages
   O all throughout practical : make a super super simple data-set to test commands: smthg like 9 reads: 3 to a gene, 3 to another gene, 3 to nothing in particular, and maybe among them 3 with adaper at the end, and 3 with low quality
         -> they will be able to use that small dataset to test their command line



Session1 - theory: https://sib-swiss.github.io/RNAseq-introduction-training/days/design.html

   objective :
      - discuss sequencing technologies
      - differentiate between technical and biological replicates.
      - choose an appropriate sequencing depth and number of replicates depending on your scientific question

   todo:
      X remove some "filler" slides about history of sequencing and alternative sequencing methods
      X provide clearer plan of analysis: describe "simple" path and QC "feedback loops" (eg, you detect a problem during DE, which leads you to change something about your mapping)

Session1.5 - UNIX and server : https://sib-swiss.github.io/RNAseq-introduction-training/days/server_login.html


  X better cheat-sheets
      X basic bash : https://github.com/RehanSaeed/Bash-Cheat-Sheet -- maybe a bit too much 
                     https://cheatography.com/deleted-124743/cheat-sheets/linux-basic-commands/ -- nice
      
      X bash scripts/variables : https://github.com/RehanSaeed/Bash-Cheat-Sheet#bash-script
      X SLURM

  X add something about script & variable



 O add some slides ??

 O explain better module how it function 
    - details in annex
    - better explain in talk


Session2 - QC : https://sib-swiss.github.io/RNAseq-introduction-training/days/quality_control.html

   X slides : OK-ish
   X practical : 
      - fastQC/multiqc: le plan est de leur faire faire d'abord la tache de maniere guidee sur le toy dataset.
                Deployer la meme chose sur les deux autres jeu de donnees est en tache extra.
        
                 En correction on interpretera avec eux 1 fastqc de chaque dataset 


      - fastq-dump : en annexe + petit jeu de donnees





Session3 - trimming : https://sib-swiss.github.io/RNAseq-introduction-training/days/trimming.html
   X clean up slides 
   X pb: trimming takes a lot of time and memory...
      --> only do it on a toy dataset
      --> compare the report from after and before trimming

Session4 - mapping : https://sib-swiss.github.io/RNAseq-introduction-training/days/mapping.html
   X clean slides. 
   X GFF and SAM part -> push to extra on website


   X practical: more hand holding on option? 
      - I like that they have to find their way through the manual, which promotes a healthy habit or properly reading the docs,
         but this part will occur at the end of day1, when their brain is quite fried...

   X compare to trimmed? --> added multiqc html report to download


Session5 - counting
   X remove that part, and discuss preprocess and normalizing more on DE??

Session6 - DE
   X slides : simplify 
   X put enrichment analysis in another slides set
   X give a bit more guidance in R practicals ?

Session7 - post-hoc
   O enrichment analysis. 
   O other kind ? (use log-count for further stuff?)
   O guidance for non-model organisms