# MAGsearch paper to do

## Bioinformatics
- [ ] Should I redo bioinformatics to generate MAGs? (no probably)
- [ ] write paper first!
- [ ] sourmash at some point
- [ ] get top hits and then do genome-grist, will get abundance-weighted - do this!
- [ ] Make sure that top metagenome hits were included in mapping validation
- [ ] Get rid of attempt to assemble and bin?
- [ ] could put in assembly loss paper
- [ ] could put in supplement
- [ ] could put nowhere
- [ ] could run spacegraphcats
- [ ] Make sure mapping is done to confirm presence

## References
- [ ] Add Antarctic biogeography papers?
- [ ] More sourmash / signatures papers?

## Questions?
Where to publish? or preprint?
- Gigascience (very computational)
- Genome Biology, write then post preprint (ask Christy)

MAGsearch with other 5?

Tech q's

Weird containment (low) - what does it mean? (containment means?)

Didn't find - not sequenced, low coverage, divergent (biogeo means?)

## Detailed To Do

- [ ] Rewrite abstract
- [ ] Read paper, what needs expansion?
- [ ] Get top genome hits to grist
- [ ] Organize github repo
- [ ] k = 31 enables matches ~91% ANI - check this
- [ ] Add Vanda sampling methods from chapter 2
- [ ] put in full bioinformatics methods or reference phormidium paper? then vanda scripts here?
- [ ] Put Vanda MAGs on NCBI
- [ ] Insert Vanda MAG stats from chap 2
- [ ] mapping only for selected metagenomes - change to top hits
- [ ] "containment affected by small metagenomes" rephrase to include low coverage and metagenome depth
- [ ] Split failed assembly and binning and mapping data into 2 tables
- [ ] Add discussion section about low containment
- [ ] Add what if means if something is not found discussion section