Roel van der Ploeg

Excited to share that my paper on Parallel Factor Analysis for longitudinal microbiome data has been published in mSystems! It details data rearrangement and processing to uncover microbial dynamics. Check it out: buff.ly/5jXEG52

Supervised Modelling of Longitudinal Human Milk and Infant Gut Microbiome Reveal Maternal Pre-Pregnancy BMI and Early Life Growth Interactions Maternal obesity is a key risk factor for excessive foetal growth and childhood obesity, yet its influence on human milk (HM) composition and the infant gut microbiome development remains unclear.…

My latest preprint is now available, featuring a collaboration with Danish colleagues. We used N-way Partial Least Squares models to study maternal pre-pregnancy BMI and infant growth, revealing complex interactions between mother milk composition and the infant gut microbiome. tinyurl.com/yc33snt6

I'm excited to share that my R package, parafac4microbiome, is updated to version 1.1.2 on CRAN: tinyurl.com/5p78hta3. This release features the full dataset from vanderPloeg2024 (tinyurl.com/y4j58kc8), an improved CORCONDIA, updated vignettes, and much more! #RStats

I presented my poster today at the Amsterdam Microbiome Expertise Center (AMEC) day at Artis zoo. I'm excited to share it here too. Feel free to message me with any questions! #AMEC2025 #Microbiome #DataAnalysis

parafac4microbiome: Exploratory analysis of longitudinal microbiome data using Parallel Factor Analysis Studies investigating microbial temporal dynamics are increasingly common, leveraging longitudinal designs that collect microbial abundance data across multiple time points from the same subjects.…

(8/8) I hope that this was helpful. If you got this far, thank you very much for your interest. Feel free to DM me if you have any questions. Also check out my paper (https://buff.ly/4hDLpFw) and the parafac4microbiome R package (https://buff.ly/4hFVtOD) for more details.

(7/8) PARAFAC reveals distinct microbial growth patterns: Proteobacteria show exponential growth in case subjects, while Actinobacteria exhibit a sigmoid growth curve in controls. This method interprets time-resolved microbial differences between groups more effectively than PCA.

(6/8) In contrast, Parallel factor analysis decomposes a three-way data array into components with three loading vectors each: one for the subjects, one for the microbial relative abundances, and one for time.

(5/8) The biplot of the PCA shows that the first principal component distinguishes case and control samples by Actinobacteria abundance, while the second separates different time points. However, the time profiles of each subject group remain unclear.

(4/8) Principal Component Analysis would decompose a two-way data table like this into components with scores for the measurements and loadings for the microbial relative abundances. This does not take the longitudinal study design into account.

(3/8) The mock data contains three phyla corresponding to 25 case and 25 control subjects across 10 time points. Initially, all subjects are similar. By time point 10, case subjects show reduced Actinobacteria and increased Proteobacteria, while controls show the opposite.

(2/8) I will highlight how this helps using a a mock intervention dataset, where we have obtained count data of three phyla corresponding to 25 case and 25 control subjects across 10 time points.

(1/8) Longitudinal microbiome data is shown as a two-way table with rows for the measurements (subject-time point combinations) and columns for microbial abundances, which can be transformed into a three-way array by adding time points as a third dimension.

Here's a thread 🧵 on using the parafac4microbiome package to track microbial changes with tensors over time. Check out my paper (https://buff.ly/4hDLpFw) and the parafac4microbiome R package (https://buff.ly/4hFVtOD) for more details. #RStats

parafac4microbiome: Exploratory analysis of longitudinal microbiome data using Parallel Factor Analysis Studies investigating microbial temporal dynamics are increasingly common, leveraging longitudinal designs that collect microbial abundance data across multiple time points from the same subjects.…

My second paper on applying Parallel Factor Analysis (PARAFAC) to microbiome datasets is submitted and available on @biorxivpreprint.bsky.social : https://buff.ly/3PNae61. It features my R package, parafac4microbiome, downloadable from CRAN: https://buff.ly/4apC9CT.

#AcademicTwitter #rStats

Multi-way modelling of oral microbial dynamics and host-microbiome interactions during induced gingivitis - npj Biofilms and Microbiomes npj Biofilms and Microbiomes - Multi-way modelling of oral microbial dynamics and host-microbiome interactions during induced gingivitis

Thrilled to share my 1st paper in @natureportfolio.bsky.social npj Biofilms & Microbiomes! We show how Parallel Factor Analysis reveals microbial subcommunities & pathways in gingivitis. Thanks to my coauthors!

https://buff.ly/421t1C5

#Microbiome #AcademicTwitter #phdlife

parafac4microbiome: Parallel Factor Analysis Modelling of Longitudinal Microbiome Data Creation and selection of PARAllel FACtor Analysis (PARAFAC) models of longitudinal microbiome data. You can import your own data with our import functions or use one of the example datasets to…

I thoroughly enjoyed making an R package that create Parallel Factor Analysis (PARAFAC) models for longitudinal microbiome data analysis. Learn more about it on https://buff.ly/4a6z3mX or my preprint at https://buff.ly/3W2nQhn #Microbiome #Bioinformatics #DataScience #RStats

G.R. van der Ploeg - About

Happy to share that my personal website grvanderploeg.com is live. You can view my CV, publications and presentations on there if you're interested.

Happy to be on this platform!