Do you still get meaningful data from lower sequencing depths, for the sake of higher cell numbers?

Hi Oliver！ I'm glad to hear that you are interested in our work. This matter requires a discussion tailored to the specific conditions of your experiment. For instance, if your research involves the deletion of large genomic regions, such as 500 kb or 1 Mb, and examines their effects on chromatin structure and RNA, such extensive deletions are likely to significantly impact chromatin architecture. Therefore, a relatively shallow sequencing depth would be entirely feasible in these cases. However, if your study focuses on interactions between enhancers and promoters, particularly when they are in close linear proximity, high-quality single-cell chromatin structure information becomes essential, necessitating deeper sequencing or specifically enrich signals from your intrested regions.

From my perspective, in most scenarios, the fewer contacts per single cell can be compensated by increasing the number of cells analyzed. This compensation requires a comprehensive evaluation of various factors, including the difficulty of sample acquisition, the specific cell types of the tissue under study, and the efficiency of the experimental conditions, to determine the necessary sequencing depth. In our publication, the 3D structure reconstruction algorithm requires a relative high number of contacts; however, there are many other algorithms available for imputation in single-cell Hi-C data. I have not conducted downsampling tests to replicate all analyses on our data, but I am inclined to believe that even with only 100,000 to 200,000 contacts per single cell (corresponding to about one-tenth to one-fifth of the sequencing depth discussed in Supplementary Figure S1D), most analyses in the article can be reproduced using appropriate imputation methods.

If you have any specific questions or need further details about our methods or results, please feel free to ask here or contact us directly. Zhiyuan.