matthew scholz

Bioinformatics Scientist. All opinions are mine, and may or may not be shared by others.

[matthew scholz @GenIgnored]

@PartridgeCG Low DNA quantity CAN be an issue, but if you get enough DNA to make a library, I'd probably recommend metagenomics.

But don't take my word for it alone!

[matthew scholz @GenIgnored]

@PartridgeCG You can look at the infographic on the README from MetScale to get a feel for what you can do with metagenomics:

https://t.co/Ybb8XWOJjw

[matthew scholz @GenIgnored]

@PartridgeCG I mean, you can definitely run functional gene analysis/pathway analysis and see what you find.

lots of thoughts on this, and I can selfishly point to tools I've worked on to process, but that's mostly bells and whistles after you make a decision.

[matthew scholz @GenIgnored]

@PartridgeCG Downsides to metagenomics:
Sensitivity: If you have the potential for lots of host DNA, it can be a problem (imagine 50 mouse reads for every bacterial read)

[matthew scholz @GenIgnored]

@PartridgeCG Reasons I recommend shotgun metagenomics:
1) Analysis is simplified
2) Library prep is WAAAY easier
3) Lower bias (still an issue from lysis, etc)

Cost: If you're having to pay for a whole flow-cell to run 16S, it likely is more expensive to do 16S until you have > 96 samples

[matthew scholz @GenIgnored]

@WhySharksMatter Is there a gritty reboot of "West Side Story", where the Sharks are on coke, and the Jets do meth, and the journalists are just confused?

[matthew scholz @GenIgnored]

@mtanichthys @PartridgeCG It's only useful if you have a computer with more than 4 cores, really. If you do, though, it's almost always faster and easier than for loops.

[matthew scholz @GenIgnored]

@PartridgeCG @mtanichthys I WANT AN EDIT BUTTON! 3 : not 4,

::: *.fasta

Sigh.

[matthew scholz @GenIgnored]

@PartridgeCG @mtanichthys Also, parallel would do what you want faster:

parallel -j [threads] awk [command] {1} {1}.mod :::: *.fasta

[matthew scholz @GenIgnored]

@PartridgeCG Or, I guess, how are you doing it with one file, and we can figure out how to expand it.

[matthew scholz @GenIgnored]

@PartridgeCG define "modify" ?? add something to the beginning, append, change completely?

[whalefact @awhalefact]

it would be very difficult to fit two whales in a 2002 toyota camry at the same time

[matthew scholz @GenIgnored]

@subsurface_life @DuffyLab IF you can chmod, the best thing I've ever seen happen in real life was the accidental space on /lib:

> mv /lib /python/

[matthew scholz @GenIgnored]

@BriannaWu @threadreaderapp unroll, if you still exist.

[matthew scholz @GenIgnored]

@michaelhoffman If you can't figure out how to do something in a particular language, it is obviously the fault of the language, and you should go ahead and learn another one.

Related, the speed of the interpreter is far more important than the portability of the code...

[matthew scholz @GenIgnored]

@bioinformer It's proof that BIRDS AREN'T REAL, is what it is. They don't even have DNA.

[L Brian Patrick @LBrianPatrick]

I don't know to whom credit is due for this piece of genius!

[whalefact @awhalefact]

when one door close another one opens but that doesn’t matter because most whales wouldn’t fit through either one

[matthew scholz @GenIgnored]

@rob_beiko Gene Godbold.