Biostatistician @MSKBiostats; PhD @EmoryRollins @laneygradschool; BSc @CityUHongKong. 🏎 🎸(Opinions are my own)
Joined March 2011
- Tweets120
- Following581
- Followers259
- Likes544
Pinned Tweet
Excited to share our R package for microbiome data analysis, published @CellRepMethods, which makes log-ratio lasso regression more accessible for R users, supporting continuous, binary and survival (competing risks) outcomes.
github.com/vdblab/FLORAL
doi.org/10.1016/j.crmeth.202…
Another good example of utilizing the powerful {gtsummary}, {ggsurvfit} and {tidycmprsk} packages for clinical data analysis @MSKBiostats. Congratulations to @KRejeski, Jaime, @RShouval, and team!
One of the more frequent questions we’ve received since the development of the ICAHT grading system is:
What about thrombocytopenia after CAR T-cell therapy?
We’ve tackled this question head-on in our new paper out now in @BloodJournal 👇
ashpublications.org/blood/ar…
An exciting real data application of finite mixture model/latent class analysis with prediction modeling. Congratulations to the team!
🚨 Just out @NatureMedicine!
Introducing InflaMix—a point-of-care tool that identifies an #inflammatory 🔥 signature predictive of #CART failure in #lymphoma
Led by Sandeep Raj (@SSRaj017) 👏@MSKCancerCenter
nature.com/articles/s41591-0… 🧵Let’s break it down
3
205
I had a wonderful #CS2023 experience with @theClassSociety, learning all kinds of clustering techniques and presenting my latent class work. Huge thanks to Dr. Tanzy Love, @BrianFranczak and colleagues for organizing this great conference!
UMAPs are among us...
I heard that UMAPs are sus so I used @lpachter lab's supervised dimensionality reduction tool Picasso to generate a much more realistic picture of gene expression in the brain.
#singlecell #Bioinformatics
Our new method to evaluate the association between pre-defined latent classes (e.g. classes defined at baseline) and competing risks outcomes:
onlinelibrary.wiley.com/doi/…
Feature: We do not require class assignment (or hard label), which avoids biases caused by misclassification.
Say goodbye to my and @StatsXin’s precious collections, forever lost by @FedEx, including all statistics books we kept since undergrad days, and signed @F1 caps and shirts by @Max33Verstappen and #SebVettel. At least sweet memories are still with us, and we will create many more!
1
3
Here Comes the Sun by NIH Director Dr. Francis Collins youtu.be/2QHRPVwa10E via @YouTube
Very excited to publish our work about latent trajectory class analysis. We incorporate GEE into finite mixture model and propose interesting ways forming posterior membership probabilities. Computational challenges are also properly addressed.
onlinelibrary.wiley.com/doi/…
1
3
I gradually understand why professors generally prefer cooler temperature in a classroom - I cannot help sweating when giving a lab session - even via Zoom...
1
A single-cell analysis of the molecular lineage of chordate embryogenesis biorxiv.org/content/10.1101/…
Our manuscript on batch effect correction is published on Bioinformatics doi.org/10.1093/bioinformati…. We conducted numerous comparisons with popular benchmark methods and provided our perspective on this challenging research area. @EmoryBIOS
2
My first collaborative research paper is just published on Head and Neck. onlinelibrary.wiley.com/doi/…
1
Feel a bit confused with the recent two incidents involving Chinese scientists at @EmoryUniversity... hard to consider them as independent events
Emory's Shan Ping Yu is about to lose his office. University officials say it has nothing to do with the scientist's Chinese origins or his petition asking the president to defend the value of having an international faculty. sciencemag.org/news/2019/06/…
1
A preprint on our attempt to mitigate batch effects in RNA-seq data. It's been a nice journey working with Dr. Tianwei Yu and bioinfo group at @EmoryBIOS!
Batch Effect Correction of RNA-seq Data through Sample Distance Matrix Adjustment biorxiv.org/cgi/content/shor… #biorxiv_bioinfo