question about snmC-seq

[Dan-121]

Hi, Zhijie: Your work is excellent but I have some questions about snmC-seq: The data you analyzed From Luo Lab offers a lot of preprocessing data, such as gene_level_human.txt.gz mCG_genebody_human.txt.gz methylation_bins_human.txt.gz methylation_bins_normalized_human.txt.gz .Can you tell me which one we use to analyze? Thank you.

[Jeff1995]

Hi Dan,

Thanks for your interest in GLUE! We used both "gene_level_mouse.txt.gz" and "mCG_genebody_mouse.txt.gz", the former for mCH level quantification, the latter for mCG level quantification. You may find the details in this data collection notebook.

Let me know if there were further issues.

Best, Zhi-Jie

[Dan-121]

Thank you for your prompt response! I am grateful to report that the issue has been successfully resolved, and I must say, the tool proved to be incredibly efficient and user-friendly. Once again, thank you for your assistance!

Dan

[Dan-121]

Please excuse me for interrupting again. I have another question: is there any difference between using genome non-overlapping 100Kb bins and using genome non-overlapping 5Kb bins when integrating data with GLUE? I would appreciate any insights you can provide. Thank you

[Jeff1995]

Ideally speaking, smaller bins should produce higher resolution both in terms of cell integration and regulatory inference, but with the following two caveats:

1. Smaller bins would produce far more nodes (20 times 5kb bins compared to 100kb) in the guidance graph, which would definitely have an impact on integration speed;
2. Smaller bins have lower single-cell detectability compared to larger bins, so there might be higher "dropout" rate and lower ability to discern cell states/types.

I'd recommend checking these two aspects before deciding on the bin size.

[Dan-121]

Thanks a lot!