Lightnews — Scholar-powered news

Will Townes @willtownes.bsky.social · 9d

Are you interested in concentration inequalities? Curious about whether there is a middle ground between subexponential and subgaussian bounds? Want to know how this all interacts with exponential tilting? Then check out my new paper recently published in TMLR: openreview.net/forum?id=BQB...

Exponential tilting of subweibull distributions

The class of subweibull distributions has recently been shown to generalize the important properties of subexponential and subgaussian random variables. We describe alternative characterizations of...

openreview.net

Will Townes @willtownes.bsky.social · Aug 26

Thoughts on peer review: willtownes.github.io/blog/2025/08...

research profile and technical blog

willtownes.github.io

Reposted by Will Townes

James Munday @jdmunday.bsky.social · Jun 9

We have a new preprint 📝 *Post-sampling degradation of viral RNA in wastewater impacts the quality of PCR-based concentration estimates* Now on bioRxiv doi.org/10.1101/2025... with @tanjastadler.bsky.social, trj2.bsky.social, Christoph Ort and others

Post-sampling degradation of viral RNA in wastewater impacts the quality of PCR-based concentration estimates

Successful wastewater-based infectious disease surveillance programs depend on regular, reliable molecular detection of nucleic acids in municipal wastewater systems. This process is challenged by the...

doi.org

1 3 9

Will Townes @willtownes.bsky.social · Jul 21

Anyone else having trouble connecting to github due to a "potential security issue"?

1

Will Townes @willtownes.bsky.social · Jul 1

Benchmarking study of single-cell feature selection finds deviance to be an effective criterion: doi.org/10.1093/bib/... . Try it yourself with our ( @kstreet.bsky.social @stephaniehicks.bsky.social @davi1893.bsky.social ) scry package @bioconductor.bsky.social : doi.org/doi:10.18129... #genomics

Characterizing efficient feature selection for single-cell expression analysis

Abstract. Unsupervised feature selection is a critical step for efficient and accurate analysis of single-cell RNA-seq data. Previous benchmarks used two d

doi.org

2 12

Will Townes @willtownes.bsky.social · Jun 3

Currently teaching data science to CMU students in beautiful Monteverde, Costa Rica! Students are working with real data from community partners: eg, permitting for new construction, flow rates from springs that supply drinking water, bell bird calls, and macroinvertebrate water quality indicators

Will Townes @willtownes.bsky.social · Mar 14

Overview of infectious disease forecasting and our Delphi research group @cmu.edu by Roni Rosenfeld: boxcast.tv/view/opening... (start at 20 min mark) delphi.cmu.edu

Opening Session AMIA 2025 Informatics Summit

boxcast.tv

1

Will Townes @willtownes.bsky.social · Jan 31

@epigenci.bsky.social makes a pretty convincing case for pseudobulk (as opposed to GLMMs) in multi-sample single cell studies here: doi.org/10.1093/bioi... . #bioinformatics

Pseudobulk with proper offsets has the same statistical properties as generalized linear mixed models in single-cell case-control studies

AbstractMotivation. Generalized linear mixed models (GLMMs), such as the negative-binomial or Poisson linear mixed model, are widely applied to single-cell

doi.org

2 10

Will Townes @willtownes.bsky.social · Jan 28

Very proud of CMU Statistics & Data Science PhD candidate Gabrielle Thivierge who just passed her thesis proposal! "Methods for performing and evaluating infectious disease forecasting and nowcasting". Check out her first paper here: arxiv.org/abs/2408.12722

Does Spatial Information Improve Influenza Forecasting?

Seasonal influenza forecasting is critical for public health and individual decision making. We investigate whether the inclusion of data about influenza activity in neighboring states can improve poi...

arxiv.org

1

Will Townes @willtownes.bsky.social · Nov 1

Cool #Rstats package for wastewater-based #epidemiology adrian-lison.github.io/EpiSewer/ #infectiousdisease

Estimate Epidemiological Parameters from Wastewater Measurements

What the package does (one paragraph).

adrian-lison.github.io

1 3

Will Townes @willtownes.bsky.social · Aug 29

Spatiotemporal variability of the pepper mild mottle virus (PMMoV) biomarker in #wastewater : a new preprint with CMU Stats PhD student Anna Rosengart and @wastewaterscan.bsky.social
collaborators arxiv.org/abs/2408.12012

Spatio-Temporal Variability of the Pepper Mild Mottle Virus...

Since the start of the coronavirus-19 pandemic, the use of wastewater-based epidemiology (WBE) for disease surveillance has increased throughout the world. Because wastewater measurements are...

arxiv.org

1

Will Townes @willtownes.bsky.social · Jul 29

Mixed Poisson distributions are useful in modeling count data (eg negative binomial is Poisson mixed with gamma). But did you know the mixing distribution can actually include negative values? Read all about it here: arxiv.org/abs/2407.17614 #probability #statistics

Mixed Poisson families with real-valued mixing distributions

Mixed Poisson distributions provide a flexible approach to the analysis of count data with overdispersion, zero inflation, or heavy tails. Since the Poisson mean must be nonnegative, the mixing...

arxiv.org

1 2

Will Townes @willtownes.bsky.social · Jul 18

New preprint: "Exponential tilting of subweibull random variables". arxiv.org/abs/2407.11386 Comments welcome! #probability #statistics

Exponential tilting of subweibull distributions

The class of subweibull distributions has recently been shown to generalize the important properties of subexponential and subgaussian random variables. We describe alternative characterizations...

arxiv.org

Will Townes @willtownes.bsky.social · Mar 18

If I am interested in prediction intervals and am comfortable with a parametric model, what are the pros and cons of conformal prediction as opposed to bootstrapping?

1 3