Clarify datasets

[shntnu]

No description provided.

[shntnu]

• delete 1.run-workflows/profiles/NCP_PROGENITORS_1/BR_NCP_PROGENITORS_1.csv.gz because it is a duplicate

df <- read_csv("1.run-workflows/profiles/NCP_PROGENITORS_1/BR_NCP_PROGENITORS_1/BR_NCP_PROGENITORS_1.csv.gz")
df2 <- read_csv("1.run-workflows/profiles/NCP_PROGENITORS_1/BR_NCP_PROGENITORS_1.csv.gz")
compare::compare(df, df2)
# TRUE

• delete 1.run-workflows/profiles/NCP_PROGENITORS_1_BRANCHING/BR_NCP_PROGENITORS_1.csv.gz because it is a duplicate

df <- read_csv("1.run-workflows/profiles/NCP_PROGENITORS_1_BRANCHING/BR_NCP_PROGENITORS_1/BR_NCP_PROGENITORS_1.csv.gz")
df2 <- read_csv("1.run-workflows/profiles/NCP_PROGENITORS_1_BRANCHING/BR_NCP_PROGENITORS_1.csv.gz")
compare::compare(df, df2)
# TRUE

df <- read_csv("1.run-workflows/profiles/NCP_PROGENITORS_1/BR_NCP_PROGENITORS_1/BR_NCP_PROGENITORS_1.csv.gz")
df %>% select(-matches("Metadata_")) %>% dim()
# [1]  380 4293

df <- read_csv("1.run-workflows/profiles/NCP_PROGENITORS_1_BRANCHING/BR_NCP_PROGENITORS_1/BR_NCP_PROGENITORS_1.csv.gz")
df %>% select(-matches("Metadata_")) %>% dim()
# [1] 380  23

[shntnu]

@yhan8 Have a look at the README

The remaining puzzle (for now) is to figure out whether the repeat progenitor plate (BR00127194) had branching feature include in the profiles or not. Can you figure that out and update the table?

#10 (comment)

[yhan8]

Can either @shntnu or @mtegtmey help confirm my understanding of the profiles below is correct.

For stem cells, the profile is located here. I am using the normalized_variable_selected profile, which if I am correct, this file has gone through normalization and feature selection. There are no branching features.

---

For progenitor cells, the profile of morphological features is located here. Please note that the csv.gz file has 4200+ features, which indicates it has not gone through feature selection process. I am going to perform a default feature selection on this profile using pycytominer. However, before I do so, is this file normalized at all? Can someone confirm?

There are 20+ branching features for the progenitor cells located here. I will add these 20+ branching features to the feature selected morphological features explained in the above paragraph to generate the final progenitor profile for downstream analysis.