Seeking Guidance on How to Annotate Unusual Cell Populations in Single-Cell Data

Question

[Daddy]

Hi Biostars Community,

I am new to the field and working with integrated single-cell transcriptome data from 26 different subtype breast cancer patients. I have identified a population that is CD45+ CD3G- CD4+, which is intriguing as CD3G- typically indicates non-T cells, yet CD4+ is a marker for T helper cells. This unexpected combination has left me uncertain about how to classify this population.

I am interested in learning about strategies or methodologies that can be employed to accurately identify and characterize such unusual cell populations. Specifically:

1. What are the best practices for confirming the identity of a cell population that does not fit conventional marker expression profiles?
2. Are there any recommended computational or experimental approaches to investigate these cells further to understand their biological significance?
3. How should one address potential technical factors such as preprocessing issues, sequencing depth, or sample preparation that might influence the identification of such cell populations?

P.S. Batch effect corrections were conducted using scVI-Tools, adhering to current best practices.

Resolved the issue: myeloid cells, including macrophages also express CD4. I am just dumb

[fracarb8]

You should check additional markers, as CD4 is also expressed by myeloid cells, and sometimes T-CD4 are better identified by the absence of CD8 than by the presence of CD4.

[Daddy]

Thank you for your insight. Just seconds after posting, it clicked that myeloid cells, including macrophages, indeed express CD4. It seems I am still on the steep part of the learning curve in the single-cell universe! I will keep this post up as a testament to the learning process and a potentially helpful marker for others on a similar journey. Appreciating the guidance in this community.