Lesson Objectives

[sstevens2]

Reply with learning objectives (and challenges to test them, if possible).

[sstevens2]

And tools

[sstevens2]

• Learners should know why we chose these tools and how to find and evaluate other tools to do similar processes

[sstevens2]

Tools:

• Biobakery?
• Mapping software to use?

[Sailendharan]

https://www.nature.com/articles/nature13568 - manuscript for reference

[Sailendharan]

Tools:

• Metaphlan2

• Humann2

• Anvio?

[Sailendharan]

https://github.com/knights-lab/SHOGUN - Shogun pipeline https://msystems.asm.org/content/3/6/e00069-18 - Shallow shotgun pipeline manuscript

[elizabethmcd]

Bloom's taxonomy for learning objectives: https://cft.vanderbilt.edu/guides-sub-pages/blooms-taxonomy/

Learning Objectives:

• Classify taxonomical composition of short reads
• Assign gene functions to short reads
• Compare enriched species/functions between patient and control groups

Analyses:

• Species identification: MetaHIT, metaPhlan (however it's spelled), Kraken from Mick Watson, anvi'o uses Centrifuge I believe. The choice of database might be different than in the paper, because many de novo bins from large-scale human gut microbiome metagenomic surveys have been published, such as: https://www.sciencedirect.com/science/article/pii/S0092867419300017
• Functional enrichment: the paper uses KEGG and eggNOG. I'm not sure if they uploaded their datasets to the BlastKOALA and GhostKOALA servers, but now the KEGG database has KofamKOALA for HMM searching, and that's a local installation. I've never used eggNOG. I'm also not sure how to set up the files for functional annotation - do they need to be clustered into gene groups first to create a non-redundant set?
• Some sort of analysis: This is where I think we need to talk more. Are we just going to show them how to make taxonomical and functional assignments, and then leave all the downstream analyses and statistical stuff to them? Because the rest of the paper's analyses for functional enrichment comparisons would probably involve some amount of R. We might want to ask questions in 2 weeks if most of the knowledge gaps are centered around the data wrangling/processing, or the actual statistical analyses, or a combination of the two.

[Sailendharan]

resource of interest from Edna - http://rpubs.com/ednachiang/MetaG_Pipeline

[Sailendharan]

Workshop objectives for Sailen's BIotech 16S workshop: Define the objectives of microbiome projects Process raw data files using appropriate quality control Run standard pipelines for marker-gene Analyze results using statistical and network approaches Recognize the technical limitations of microbiome studies