search

igrabski / sc-SHC

Significance analysis for clustering single-cell RNA-sequencing data
87 stars 10 forks source link

Cannot extract p-values from cluster comparisons #22

Closed sg3451 closed 2 months ago

sg3451 commented 2 months ago

Is there a way to extract the p-values from the 'testClusters' function? The output is currently a list with 2 elements but none of them contain statistical significance results. One element of the list appears to be a dendogram but it is hard to interpret it.

igrabski commented 2 months ago

Hi, the p-values are printed after the colon next to the node/cluster labels in the dendrogram-type output (the second element of the list). They can be interpreted as the p-value corresponding to splitting that node/cluster. If no p-value is printed next to any particular cluster, it's because no further split was evaluated within that cluster (e.g. no further split may be possible if that cluster already represents one of the original inputted clusters).

sg3451 commented 1 month ago

Thanks very much for the prompt response! Just so I understand it correctly, in the following output, the splits at Node 0, Node 1, Node 2 and Node 3 are significant. Ultimately, scSHC is reporting 5 clusters based on the node splits (named Cluster 1-5) - is this correct? There are no further splits from any of these 5 clusters?

Node 0: 0 2 ¦--Node 1: 0 3 ¦ ¦--Cluster 1 4 ¦ °--Node 3: 0 5 ¦ ¦--Cluster 4 6 ¦ °--Cluster 5 7 °--Node 2: 0 8 ¦--Cluster 2 9 °--Cluster 3

On Wed, May 29, 2024 at 4:03 PM igrabski @.***> wrote:

Hi, the p-values are printed after the colon next to the node/cluster labels in the dendrogram-type output (the second element of the list). They can be interpreted as the p-value corresponding to splitting that node/cluster. If no p-value is printed next to any particular cluster, it's because no further split was evaluated within that cluster (e.g. no further split may be possible if that cluster already represents one of the original inputted clusters).

— Reply to this email directly, view it on GitHub https://github.com/igrabski/sc-SHC/issues/22#issuecomment-2138260272, or unsubscribe https://github.com/notifications/unsubscribe-auth/AR4XNDP5BDPFLMU3KUETNF3ZEY7CLAVCNFSM6AAAAABIPVQVDCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDCMZYGI3DAMRXGI . You are receiving this because you authored the thread.Message ID: @.***>

igrabski commented 1 month ago

Yes that is correct!

sg3451 commented 1 month ago

Excellent - thank you again for the quick response. I appreciate it.

Regards,

Sujoy

On Thu, May 30, 2024 at 10:14 AM igrabski @.***> wrote:

Yes that is correct!

— Reply to this email directly, view it on GitHub https://github.com/igrabski/sc-SHC/issues/22#issuecomment-2139887164, or unsubscribe https://github.com/notifications/unsubscribe-auth/AR4XNDNAUBSWBEQBTLPPYPLZE465HAVCNFSM6AAAAABIPVQVDCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDCMZZHA4DOMJWGQ . You are receiving this because you authored the thread.Message ID: @.***>