Cell type mapping

[liyawang]

1. Given a marker from one cell type, Astrocytes
2. Calculate the mean for each cluster (C1, C2, C3, C4, C5), then rank them as S1, S2, S3, S4, S5
3. Compare S1 with S2 (z test) if significant, assign S1 (C3) to Astrocytes if not, merge S1 with S2, compare with S3 if significant, assign S1 and S2 to Astrocytes if not, give up
4. Go back to 1 for a different marker to generate an assignment table Asctrocyte: C3, C3/C4, C4 --> C3/C4
5. Go to 1 for next cell type

[liyawang]

1. Given a marker from one cell type, Astrocytes
2. Calculate the mean for each cluster (C1, C2, C3, C4, C5), then rank them as S1, S2, S3, S4, S5
3. Compare S1 with S2 (t test), assuming p-val < 0.000000001 stats.ttest_ind(adata[adata.obs["clusters"] == "C3", "Pdgfrb"].X, adata[adata.obs["clusters"] == "C4", "Pdgfrb"].X, alternative="greater", trim=0.3) if t-statistic > 20, assign S1 (C3) to Astrocytes else, S1 undefined
4. Go back to 1 for a different marker to generate an assignment table Astrocyte: C3, C4, C4 --> C4
5. Go to 1 for the next cell type