Hi everyone,
I’m a high school student developing a computational research project in pancreatic ductal adenocarcinoma (PDAC), and I’m looking for someone currently working in bioinformatics, computational biology, or single-cell analysis who would be willing to briefly review my methodology.
The study uses publicly available PDAC single-cell RNA-seq data and involves RNA velocity/scVelo, CellRank-based fate inference, pseudotime, and CytoTRACE. I have developed a composite index intended to quantify variation in inferred terminal-state probability distributions at the individual malignant-cell level.
I’ve already developed the computational workflow, research questions, hypotheses, and planned statistical analyses. I’m mainly looking for a technical review of the methodology: whether the workflow and statistical tests are appropriate, whether the assumptions are sound, and whether there are important confounders or methodological issues I may have overlooked.
I have a qualified scientist familiar with the broader research area, but they are currently quite busy, so I’m hoping to get an additional perspective from someone who works directly with these types of computational analyses.
I’m happy to provide the full methodology privately rather than posting the complete research framework publicly. Any level of technical feedback would be greatly appreciated; you may reach out to me through email: matteogalgo.ph@gmail.com.
1 answer
Dear Matteo, I came across your post on BioStars regarding your single-cell RNA-seq project on Pancreatic Ductal Adenocarcinoma (PDAC) and would be glad to offer a technical review of your computational methodology.
I hold an MSc in Applied Bioinformatics from Aristotle University of Thessaloniki and a BSc in Statistics & Financial Mathematics from the University of the Aegean. My background centers on high-throughput sequencing data workflows, trajectory/fate inference, and statistical modeling in R and Python.
I would be happy to review: Workflow Integrity & Assumptions: The sequencing of your preprocessing steps, velocity estimation using scVelo (e.g., evaluating dynamic vs. deterministic models and spliced/unspliced ratios), and terminal-state inference via CellRank. Composite Index Formulation: The mathematical formulation and assumptions of your composite index for quantifying terminal-state probability distributions at the single-cell level. Statistical Rigor & Confounders: Assessment of normalization artifacts, batch effects, cell-cycle confounding, pseudotime trajectory stability, and the appropriateness of your planned hypothesis testing.
Please feel free to send over the documented methodology, workflow outline, and planned statistical tests whenever convenient.
Best regards,
Dimitris GiouvanisMSc
Bioinformatician | BSc Statistician Thessaloniki, Greece
Email: giouvanis.16017@gmail.com
LinkedIn: linkedin.com/in/dimitris-giouvanis | GitHub: github.com/Giouvanis
Log in to answer this question.
In general it would be best to find people who are near you and with whom you can meet in person to go over things.
Biostars has no facility for direct messaging, so you may want to add some kind of contact (email or LinkedIn page to the post above) taking appropriate precautions.