2023-02-14

This is PMAS's COMMONS Lab daily Open Notebook.

Today is 2023.02.14

Todo today

Set a meeting with people.marco-visani (Private)

Work on the biostats_toolbox

graphml generation step. Find a way to merge a df (or several) to a graph object in a smoother way

Doing

Qual comparison

mapp_project_00020 mapp_batch_00036 mapp_01_39_x

mapp_project_00020.mapp_batch_00036 (Private)

Updating DBGI program

communication.meetings.2023.02.22.program (Private)

Paused

Done

Notes

The SPHN (Swiss Personalized Health Network)

Have been funded for 135M CHF

Semantics standards not common in healthcare

Standards don't send you money now, they send you money later

G. Grieve

Q: Hi Katrin. Thanks for the talk ! On one slide you mentioned the importance of efforts regarding integration of omics data within the SPHN KG. Are these future tasks or are they currently undertaken ? I am particularly interested to know wether they are concrete things done on the integration of metabolomics data. What are the challenges here according to you ?

National Data streams.

Integration of genomics + other omics Two of them integrate metabolomics data Partners on the data stream. Curated with an expert team in Zurich

Who are these teams ?

Interoperability framework. Sabine Oesterle Led by https://www.sib.swiss/directory?cn=sabine%20oesterle

Visualapps

Gerhard Kober gerhard.kober@visuapps.com

FIHR Stores http://hl7.org/fhir/ RuleML https://en.wikipedia.org/wiki/RuleML

Privacy preserving dashboards

Varsha Gouthamchand Maastro (ND)

Triplifier (JS based) converts and relational DB to RDF

Problems of AI in Drug Discovery

BenchSci

Check how to setup granular autorization on a git organizations

Discussion with Marco

Why should molecules be normaly distributed ? Have a look why object in Nature are usually normally distributed ? Are they exceptions ?

near field NMR ? https://academic.oup.com/gigascience/article/doi/10.1093/gigascience/giac124/6980761 https://github.com/mandelbrot-project/pf_1600_datanote

todo PM :

prepare the 1600 pf dataset
prepare DBGI

MS paper Marco

ENPKG

have the 1600 pf dataset ind files uploaded on Zenodo
Think about a structure for individual "capsule" per profile. It would include all required metdata, raw files and resulting individual RDF Could be uploaded to Zenop using their API. Each profiled sample wpould thus benefit from its individual doi

Q: what wouldn't fit in such a capsule ?

Todo tomorrow, another day ... or never !

Mechanism to copy assets

When transfering daily notes from notebook.xxxx. to open-notebook.xxxx the assets stay in the orignal vault where they were pasted. This leads to broken link.

Today I learned that

glimpse(df)

is a great way to peak into a df

Added to R