brianjohnhaas
commented
5 years ago Hi,
Good questions. There are some normal types that show different signatures across chromosomes, and by restricting to a single normal cell type that is not the best match to the tumor, it can bring out patterns that could lead to false interpretations. Essentially, in the tumor cells, you're looking at the residual expression after subtracting the normals, and so the choice of normals can have an impact on the results. For maximizing specificity, you'll want to use the full collection of normal cells, partitioned into their different groups.
hope this helps,
~b
On Tue, Apr 23, 2019 at 7:09 AM AMA111 notifications@github.com wrote:
Dear inferCNV developers, I am running inferCNV tool to infer CNV from multiple myeloma patients using 10x SC-RNA-seq data. In multiple myeloma, the plasma cells are the tumor cells and we would believe that the other cell types are benign cells in the bone marrow microenvironment. For such analysis, I used only one patient to infer CNV from the plasma cells compartment and testing if we use different cell type as a reference, could be that by doing so the CNV inference would change and this could be reflecting true biology or due to technical reasons ( the inference Algorithm, number of cells on the reference, sequencing issues, ...).
From the first glimpse, I can observe a strong signal by using any of the references cell types. However, I can observe a unique strong deletion signal in (chrom 1, 6 and 12) and a stronger deletion signal in (chrom 16 and 17) by using neutrophils as a reference in comparison to using T-cells as a reference.
[image: Patient_4_inferCNV] https://user-images.githubusercontent.com/29317258/56576252-25969180-65c8-11e9-96b9-f69275e9246f.png
How can I interpret these observations by using different cell types as a reference? Your comments or suggestions would be appreciated.
Best, Abdelrahman
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/broadinstitute/inferCNV_examples/issues/2, or mute the thread https://github.com/notifications/unsubscribe-auth/ABZRKX7RDSTCC6XSGDHOA6TPR3U6FANCNFSM4HHW5DJQ .
--
Brian J. Haas The Broad Institute http://broadinstitute.org/~bhaas http://broad.mit.edu/~bhaas
AMA111
Dear inferCNV developers, I am running inferCNV tool to infer CNV from multiple myeloma patients using 10x SC-RNA-seq data. In multiple myeloma, the plasma cells are the tumor cells and we would believe that the other cell types are benign cells in the bone marrow microenvironment. For such analysis, I used only one patient to infer CNV from the plasma cells compartment and testing if we use different cell type as a reference, could be that by doing so the CNV inference would change and this could be reflecting true biology or due to technical reasons ( the inference Algorithm, number of cells on the reference, sequencing issues, ...).
From the first glimpse, I can observe a strong signal by using any of the references cell types. However, I can observe a unique strong deletion signal in (chrom 1, 6 and 12) and a stronger deletion signal in (chrom 16 and 17) by using neutrophils as a reference in comparison to using T-cells as a reference.
How can I interpret these observations by using different cell types as a reference? Your comments or suggestions would be appreciated.
Best, Abdelrahman