Issues with using hicTransform

[HengkuanLi]

Hi, I have some questions regarding the use of hicTransform, and I would greatly appreciate your help. Thank you very much.

1. When generating the O/E matrix, should I use the raw counts or the normalized data (e.g., ICE, KR)? Additionally, can I use the data processed by hicNormalize to generate the O/E matrix? Or is it also acceptable to use the matrix processed by both hicNormalize and hicCorrectMatrix?

2. To generate the Pearson matrix, is it necessary to first generate the O/E matrix and then use the O/E matrix to further generate the Pearson matrix? I noticed you mentioned this in a previous answer, but the current documentation does not include this information. I would like to confirm whether this step is still required—i.e., first generating the O/E matrix and then using it to derive the Pearson matrix.

3. When generating the O/E matrix, which approach is more reasonable: processing the entire genome at once or processing individual chromosomes separately?