New subgroups results after batch correction for each of the scales

[eathieniti]

1. Plot heatmaps to describe the clusters
2. Plot trajectories
3. Create networks of main molecules that differentiate the subgroups: either they are associated with higher progression or they MARK the patients that will have worse progression

[eathieniti]

Batch effect correction details: https://github.com/eathieniti/mofa_tutorial/issues/2

Relationship with neutrophils and lymphocytes ? https://github.com/eathieniti/mofa_tutorial/issues/5

Clusters on the factors used to create them

Median scores for clinical scales for the MOFA subgroups:

Clinical score over time

\todo Question to discuss: How many clusters 3 or 4?

[eathieniti]

Molecules clusters centres = 3

DE Comparisons per subgroup

Question why: does group 3 have the most DE genes? Is it a different age? Shall we age-match?

[eathieniti]

MOLECULAR TRAJECTORIES

TODO:

1. fix for the blood samples cell counts
2. deseq for mirnas and save them too
3. Input into mirna-rna network
   4. Time omics to get the trajectories for each of the subgroups

[eathieniti]

Compare enrichment paths

[eathieniti]

TODO : before closing rerun clusters with neutrophil score corections

[eathieniti]

New subgroup analysis with neutrophil corections

Covariates of 4, 9?

Clinical score values of NP2PTOT are associated with Neutrophil counts and neutrophil to lymphocyte ratios