PGP public key
Fingerprint B829 F851 00B7 9CB7 6897 DCA7 A881 5DA8 5150 5ABF
Bioinformatics software
Topconfects: Bioconductor R package. Top differential expression by confident effect size. A more intuitive way to use "TREAT".
Weitrix: Bioconductor R package. Tools for working with a matrix of data measured with varying precision or with missing values.
Langevitour: Javascript with R wrapper. Interactively tour a high-dimensional numeric dataset at 60 FPS. Uses small motions to convey which points are near each other.
Varistran: R package. Variance stabilizing transformation for RNA-Seq data. This is a useful pre-processing step prior to visualization, and also makes clustering and statistical analysis more straightforward. Includes a shiny app for assessing RNA-seq data quality with heatmap and biplot.
Tail-Tools: Analysis pipeline for PAT-Seq data.
Nesoni: (No longer maintained.) Tools for processing bacterial high-throughput sequencing data.
Plasmodium Autocount and Semiautocount: Cell counting on blood smears.
Talks and slides
Weitrix
2020-10-12 Weitrix: calibrate precision weights for diverse types of data in a matrix, explore and test weighted or sparse data (for Bioconductor Asia 2020).]
2020-11-04 Video
Topconfects
2019-08-21 An introduction for a general audience (for the NUMBATS group).
2020-03-10 A more technical presentation (WEHI Bioinformatics Seminar).
Machine learning
2017-06-14 Introduction to logistic regression and decision trees, and a description of my Melbourne Datathon 2017 Kaggle entry
Tiling patterns
2022-01-27 Video about mathematical weirdness in tiling patterns (slides).
Writing
Publications on Google Scholar.
PhD thesis (2005) on image texture synthesis. I also have several publications relating to my PhD work on texture.
Introduction to R, version 2 (2018) (version 1 here) Workshop material introducing R.
Programming and tidy data analysis in R (2019) (based on an ealier workshop R more (2016)). Workshop to go from beginner to intermediate R usage, which I have developed and presented with other members of the Monash Bioinformatics Platform and the Monash Data Fluency initiative.
Linear models in R (2018) Workshop material introducing linear models in R. Linear models are a very broadly useful statistical tool, and also necessary background knowledge for using Bioconductor packages such as limma.
Dance of the CIs, a Javascript app version of the "Dance of the CIs" from the "New Statistics". Illustrates the meaning of confidence intervals, standard error, standard deviation, and prediction intervals.