Meeting 2024.06.20 - Githubissues

immunoteam / ukb_tumor

1 stars 0 forks source link

Meeting 2024.06.20 #12

Open KBalazs1987 opened 2 weeks ago

KBalazs1987 commented 2 weeks ago

Survival analysis (Cox regression and ggforest plot)

25 different tumors and no cancers
PTV burden as a continuous variable
Variables in the Cox models: sex, ethnic group, PCA1-10, PTV burden
some exclusions (low nb of patients in any sex, patients in rare ethnics)

KBalazs1987 commented 2 weeks ago

PTV burden cutoffs:

based on all patients (1: q20, q40, q60, q80, 2: q10, q20, q80, q90)
based on patients with cancer (1: q20, q40, q60, q80, 2: q10, q20, q80, q90)
based on patients with a given tumor (1. q20, q40, q60, q80. 2: q5, q10, q15, q20, q25, q75, q80, q85, q90, q95. 3. q10, q20, q80, q90)

Cancers are classified into 8 groups based on histological type

BagiLaura00 commented 2 weeks ago

This is the plot for the survival analysis: update_ptv_forest_surv

The incidence analysis has the same variables as the survival analysis. The only difference is that in this case the low nb of the patiens in any sex was not a problem. Incidence analysis from birth: update_ptv_forest

Incidence analysis from first attending: update_ptv_forest_a