Fw: Inquiry About Optimal Use of methylKit for FFPE Samples & Meeting Request

4 views
Skip to first unread message

Zeinab Abdelrahman

unread,
Oct 22, 2025, 8:03:13 AMOct 22
to methylkit_...@googlegroups.com
Dear team,
I hope this message finds you well.
My name is Zeinab, a research fellow at the Centre for Public Health, Queen's University Belfast. I’m currently working with Bisulfite sequencing data derived from FFPE samples, and I’m using the methylKit R package for methylation analysis. As you may know, FFPE-derived data often contains a high proportion of missing CpG values (NAs), which poses challenges for downstream analysis and object reconstruction within methylKit.
I would greatly appreciate your guidance on the optimal use of methylKit in this context — particularly regarding any recommended preprocessing steps or parameter settings that improve robustness for FFPE samples.
Additionally, I’d be very grateful if we could arrange a brief meeting to discuss this further. I believe your insights would be invaluable in helping us refine our analysis pipeline.
Please let me know if you’d be available for a short meeting at your convenience.
Best regards,
Zeinab Abdelrahman

________________________________________________________________________


Zeinab Abdelrahman, PhD

Research Fellow  

Centre for Public Health - Molecular Epidemiology 

Institute of Clinical Sciences - Block A

Royal Victoria Hospital

Queen's University Belfast





From: Zeinab Abdelrahman <z.abde...@qub.ac.uk>
Sent: Wednesday, October 22, 2025 12:53
To: Altuna...@mdc-berlin.de <Altuna...@mdc-berlin.de>
Subject: Inquiry About Optimal Use of methylKit for FFPE Samples & Meeting Request

Dear Altuna,
I hope this message finds you well.
My name is Zeinab, a research fellow at the Centre for Public Health, Queen's University Belfast. I’m currently working with Bisulfite sequencing data derived from FFPE samples, and I’m using the methylKit R package for methylation analysis. As you may know, FFPE-derived data often contains a high proportion of missing CpG values (NAs), which poses challenges for downstream analysis and object reconstruction within methylKit.
I would greatly appreciate your guidance on the optimal use of methylKit in this context — particularly regarding any recommended preprocessing steps or parameter settings that improve robustness for FFPE samples.
Additionally, I’d be very grateful if we could arrange a brief meeting to discuss this further. I believe your insights would be invaluable in helping us refine our analysis pipeline.
Please let me know if you’d be available for a short meeting at your convenience.
Best regards,
Zeinab Abdelrahman

________________________________________________________________________


Zeinab Abdelrahman, PhD

Research Fellow  

Centre for Public Health - Molecular Epidemiology 

Institute of Clinical Sciences - Block A

Royal Victoria Hospital

Queen's University Belfast


Reply all
Reply to author
Forward
0 new messages