Hi everyone, I'm working on a single cell dataset for my first time, i used Seurat for the analysis of the data quality control and clustering. At the end of the clusterization process i have the situation which you can watch in the plot that come with this post. My question regards the cluster 8. In the analysis after clusterization this cluster show, using findAllMarkers, the higher p-value for the "top marker" (4.746591e-18) while all other clusters have values not higher than (4.492991e-153), and also if we plot the three metrics nFeature nCount pct.mt find a very low values for the first two in cluster 8. the question is, how it's the best management of such data given what we can see? it seems to me that the cells grouped in cluster 8 are not very meaningful, so i don't want to carry on something which can create confounding analysis later but i also think that approach like remove the cluster could be too drastic. so i try to find some counsueling for that. If more information is needed teel me and i try to be more clear!
https://postimg.cc/gxcRpvSB][img]https://i.postimg.cc/gxcRpvSB/Schermata-2021-01-31-alle-23-06-29.png
https://postimg.cc/68HZpyF6][img]https://i.postimg.cc/68HZpyF6/Schermata-2021-01-31-alle-23-14-01.png
What single cell platform is this? At least some of these look like empty droplets with some ambient RNA contamination to me. A higher/lower p-value for a cluster's markers does not relate to the overall quality of the cell.
If you believe they are not actually cells, they can be removed.
it's 10X, i understand what you mean i also tried to treat my data with soupX to remove ambient RNA contamination but this cluster seems to be not affected from the method application.