gladstone-institutes/bioinformatics-workshops

Workshops presented by the Gladstone Bioinformatics Core

Main Page

Visit https://github.com/gladstone-institutes/Bioinformatics-Workshops/wiki for a curated set of user tutorials.

Accessing Presentations and Materials

Presentations are organized into directories, containing:

• Presentation in powerpoint format for downloading
• Presentation in pdf format for viewing directly on GitHub
• Necessary data files

bioinformatics-workshops's Issues

Improvements for Intro to R (April 2023)

• Add more examples for each of the basic principles
• Remove the reading in the data part from Part 1
• Add more live coding in part 1
• Add more explanation of the PanTHERIA dataset

Pre-workshop instructions:

• Install tidyverse before the workshop
• Give them an Rscript to run that checks their R version and tidyverse installation (Include screenshots of Rstudio with the instructions)

Overall emphasize the importance of reaching out to us if they have installation issues in the pre-workshop instructions/email.

Related to #3

sc-RNAseq workshop (January 2023)

• Demo and concepts together
• Rethink some of Khrisna's slides (Michela)

Intro to Pathway Analysis (Sept 2022) - Ideas for next workshop

• Consider skipping the instruction and hands-on part for data visualization, instead simply showing several (real) examples.
• If we keep data viz, the workshop should be split into two days
• Presentation: Include a lot more general background including about the Gene Ontology, pathway databases.
• Presentation: Include some examples of pathway analysis from literature
• Presentation: reformat the section about drawbacks/caveats to be more clear. Maybe needs more background on how permutations are used etc.

Improvements to Intro to Pathway Analysis (for Oct 2023)

The survey feedback is very limited but from what I recall (and one comment in survey) the hands-on session was the only real issue. Some changes to consider:

• Provide more example datasets to choose from for the online tools
• For each example dataset, add a short explanation of the data/question being asked to better relate to data that the researchers might have
• For pre-workshop instructions: Emphasize that its ideal for participants to bring their own data, or list of genes, for the hands-on session.

Intro to R workshop (July 2022) - ideas to consider for next workshop

1. include the initial demo about read.table in session 1 instead of session 2
2. include more breaks on second day
3. include an objective/goal for the data during demo in session 2
4. have exercises for participants to complete before the workshop (e.g. some starting learnR modules)
5. have more TAs during both sessions
6. dedicated time to solve queries in chat
7. start with smaller dataset in session 2
8. more smaller examples instead of one dataset with no clear aim
9. Explain more how session 2 demo will be helpful in day-to-day work

Intro to GWAS and PRS score

• Next time not during/after holiday time
• Test PLINK2 on windows

Intro to R Data Analysis - August 2023

• More hands on examples in part 1
• Simplify the hands on session in part 2
• Add more examples of each tidyverse function in part 2 before the hands on session

Experimental design and HT

1. Next time try to reduce to 2 session 2.5 hours each. Exp Design is first 45 min.

Pathway Analysis workshop updates 2024

• Add column to comparison table for number of libraries
• Add DAVID to list (moved MSigDB to "database-specific" bullet point since we describe it in terms of a data source more than the actual tool)
• Test webtools and document bugs before next workshop (e.g., WPIDs that fail to view in webgestalt and IEA)

scRNAseq - reduce the second session to 1.5/2 hours

Bulk RNA-seq workshop (July 2022) - ideas to consider for next workshop

1. Have 30 min at the starting of session 1 to help the participants with getting the Docker container running and logging into wynton.
2. Share the commands for logging into wynton and starting docker (e.g. docker images or docker version commands) before hand as per-workshop instructions
3. Keep session 1 for concepts by Michela and session 2 for demo with Ayushi
4. Let Wynton team know in advance that they may receive a lot of new account requests
5. Share email about creating wynton account at least 1.5 weeks before the workshop.
6. Windows commands for the corresponding mac/linux commands
7. Pre-workshop instructions to try basic commands (e.g. ls, cat, cd, etc)

Pathway Analysis workshop updates

Based on the Oct 31 session, the following updates should be considered:

• There was confusion about where in the analysis pipeline we were focusing. We could have a slide near the beginning illustrating an analysis pipeline and then clearly state that we are not covering prior steps. Others might already have a pipeline slide. Ask around. -> Added slide in the beginning: https://docs.google.com/presentation/d/1Qheb-TiSAXh0P1Wsu-3GYi-TY_5ubTJmlAWEAipgWig/edit#slide=id.g2a00b9273dc_0_7.
• In "Exploratory Analysis" slide we could contrast with "Confirmatory Analysis" to help define the terms. Might also use the term "Recommender system, like Netflix" to help give an intuitive sense. -> Added slide here: https://docs.google.com/presentation/d/1Qheb-TiSAXh0P1Wsu-3GYi-TY_5ubTJmlAWEAipgWig/edit#slide=id.g2a00b9273dc_0_39 (I consider this a draft)
• We miss one of the key strengths of pathway analysis: detecting biological effects even if individual gene changes are not statistically significant. A slide with a data overlay would probably make this clear.
• Add mSigDB and Cytoscape to lists of other commonly used tools
• Mention that ORA thresholds can be set to include up, down, or both (i.e., ignoring sign). Each of these three sets would be valid to test; each would be interpreted slightly differently.
• The session didn't have a clear end. Could have a slide called "Wrap-up" where you invite folks to continue with hands-on exercises, their own data, etc. or can take the survey and leave early. -> moved slides around, and updated this slide.

Single-cell RNA-seq workshop (Oct 2021) - ideas to consider for next workshop

Hi All,

Looking over the feedback for the single-cell RNA-seq workshop, I had a few ideas that I thought we could consider for the next time we do this workshop. I added this GitHub issue to stash these ideas in one place. Here are a few ideas from me:

1. Provide deep-dive materials, like papers and math details -- probably after the workshop?
2. Add scATAC-seq? As workshop slides or extra references.

If you have some ideas in mind, please feel free to add to the list by commenting/replying on this issue.

Thanks,
Ayushi

Intro to Linear Mixed Effects Models - April 2024

• The introduction to LMMs could have been more clear. The example that was written out by
  hand was difficult to read. It would have been nice to have the first model implemented in R be
  the exact same one as the example that was written out by hand.

• It would also have been
  helpful to have conceptual examples of fixed vs random effects (before the R implementation
  section).

• Basic introduction to statistical terms and interpretation of plots could be more descriptive.

• I think it could be helpful to break into breakout groups to step through some of tutorial.

• add some more materials for further reading

• Sample code on how to conduct this analysis with tidy (i.e. analyzing multiple chemicals with
  longitudinal measurements in an efficient manner) and how to present the results cleanly.

• Interaction terms between two fixed effects

• Applying linear mixed models to RNA-seq analysis (for example with packages dream, dreamlet, DESeq2, etc) and how to assess whether the model is working properly

• more on longitudinal data for non statistical background.

• Logistic regression!

Intro to scATAC-seq - part 1 and 2

• Double check the clarity of the polls
• Reduce background slides
• Merge Part 1 and part 2
• Organize one workshop in 1 day morning and afternoon + 1 day for integration vs 3 days 2 hours workshop
• Demo and slides back to back
• Send the recording of scRNA-seq along with the pre-workshop instructions( not possible?)

Brainstorming topics for Working on wynton

2 weeks before the workshop, coordinate with Andrew (Gladstone IT team) and the Wynton Admin team. Let them know that they might receive bulk requests for new wynton accounts in preparation for the workshop.

Working on Wynton - April 2024

Issues, feedback and ideas from 4/15/24:

• typo: plog vs plog1
• BeeGFS module was confusing and users disengaged. Takeaway messages didn't stick (too early). Suggestion: save this for "advanced tips and tricks" at end of Day 2
• Demo a small job early on (without them following along) to help illustrate nodes, scratch and storage, before those modules.
• Add cd line to Install samtools code chunk
• Add step 4 for adding samtools to path and testing installation
• Add a slide illustrating multiple log nodes connecting to muliple dev nodes, connecting to one file system to help make the point that it doesn't matter which log or dev node you ssh to, you are working with the same files and folders.
• Run same example container on dev and on compute node.
• Revise definition for container (see below)
• One downside of two-day split is the drop off (~65% retention). If keep two-day, then you really have to "sell" the second day and don't describe it as optional or "if".
• Maybe move all container modules to Day 2 in the context of "more advanced" and move a generic compute example to Day 1
• Typo: S for $ in script

Experimental design and hypothesis testing

• Reduce the experimental design part and reduce to 2 days workshop

Bulk RNA-seq data workshop improvements (from survey results)

[ ] Add a slide on publicly available data, how to access it and how to clean it up (maybe share a script for clean up?)