We are also delighted to welcome Andrew Bowerman to the ANU AAGI team 😊 We are very pleased to have Andrew on board and look forward to working together.

On the 12th of August 2026, The ANU-APPN (Australian Plant Phenomics Network) hosted a facility tour of its faculties at the ANU Research School of Biology. Our tour group was given an opening remark by Dr. Richard Poire, the director of the ANU chapter of the Australian Phenomics Network.

A notable thing mentioned in the opening remark was how rapid the network has grown. With the new strategic plan introduced in 2023, the number of nodes grew from two nodes in the University of Adelaide and Australian National University into 9 total nodes spread across the entire nation. We were also given an overview of operations at ANU-APPN and we understood that they are developing a national plant phenotyping asset that will deliver results in the field of plant science to be used by researchers.

Our tour began at the growth pods tucked behind the Forestry Building at ANU. These climate-controlled chambers allow researchers to precisely simulate lighting, oxygen levels, and other environmental conditions required for plant growth research. The facility is currently in active use by PhD students conducting cutting-edge research, with enough capacity to accommodate a wide range of plant species simultaneously.

Another notable facility we were shown is a multispectral 3d laser scanner that scans plants reconstructs a 3d scan of them. It moves over the plants and combines 3D laser triangulation with multispectral imaging to build a detailed 3D model of each plant, capturing traits like height, leaf area, and canopy structure. Because it works optically rather than through physical contact, researchers can gather this fine-grained data repeatedly over time without disturbing or damaging the plant structures. What was insightful to me is the software used to capture this information is also developed by the APPN, exhibiting the importance of software engineers to the research process.

Deeper inside the Linnaeus Way building, we were introduced to Conviron growth chambers, compact units used to maintain controlled conditions such as lighting, temperature, and humidity for plant research. Unlike the larger growth pods, these chambers are designed for smaller-scale experiments, with multiple shelves allowing several plant samples to be grown under monitored conditions at once.

We concluded the facility tour with Dr. Poire mentioning the other cutting edge research being done across other APPN nodes, how well the team is adhering to the 2023 - 2028 strategic plan, and hopes for future improvement. All-in-all, it was an insightful facility tour that broadened my horizons on plant research being conducted at ANU and across Australia.

• Postdoctoral Fellow in Statistics, and
• Research Software Engineer (no PhD qualification required).

Candidates for both positions must possess strong scientific programming skills, demonstrated experience working with real data, and a substantial commitment to reproducible and transparent workflows. We prefer candidates that are already familiar with some biology or analysis of plant improvement programs.

For the postdoctoral fellow, we are particularly interested in candidates with expertise in statistics with potential to publish in reputable journals. More detail for this job can be found here. The job will close at May 19 2026 23:55 AEST.

For the research software engineer, we prefer candidates who have developed R packages and have experience with software development best practices, including version control, testing, and documentation. More detail for this job can be found here. The job will close at May 5 2026 23:55 AEST.

What is heritability?

In genetics, heritability quantifies how much variation in a trait is due to genetic differences vs. environmental effects. It provides an indication of how evolvable traits are which is important for fields such as evolutionary ecology, quantitative genetics and breeding research.

# install.packages("heritable")

library(heritable)

# remotes::install_github("anu-aagi/heritable")

library(lme4)
fit_lme4 <- lmer(y ~ rep + (1|gen) + (1|rep), data = lettuce_phenotypes)

library(asreml)

Online License checked out Fri Aug 14 20:59:51 2026

fit_asreml <- asreml(
  fixed = y ~ rep,
  random = ~ gen,
  data = lettuce_phenotypes,
  trace = FALSE
)

H2(fit_lme4, target = "gen")

   Cullis     Oakey    Piepho     Delta  Standard 
0.7737572 0.7737572 0.7725405 0.7737572 0.7725219 

H2(fit_asreml, target = "gen")

A design matrix was not found in the asreml object. Building a design matrix.
Registered S3 method overwritten by 'car':
  method           from
  na.action.merMod lme4

   Cullis     Oakey    Piepho     Delta  Standard 
0.7737597 0.7737597 0.7725430 0.7737597 0.7738271 

On 1 December 2025, the Australian Statistical Conference 2025, held in conjunction with the Statistical Education Symposium, officially opened in Perth, Western Australia. The five-day event was jointly hosted by Curtin University and the Statistical Society of Australia (SSA), attracting statisticians and researchers from across the country to gather and exchange ideas.

Throughout the conference, several eminent scholars delivered keynote speeches, offering in-depth insights into the future directions and development trends of statistical science under the evolving landscape of technology and society. Researchers from various fields also presented their latest work on frontier topics in statistics, engaging in extensive discussions that further promoted academic innovation and cross-disciplinary collaboration.

At the closing ceremony, the conference announced this year’s honorary awards. Warren Muller, from the ACT Branch, received the Society Service Award in recognition of his longstanding and outstanding contributions to the statistical community.

The conference concluded successfully and achieved positive outcomes. Dr. Patrick Li, a postdoctoral researcher from the ANU-AAGI project team, was invited to deliver a presentation titled “Automated Assessment of Residual Plots with Computer Vision Models,” which received strong interest and positive feedback from attendees.

The organisation of the conference was made possible thanks to the efforts of the local organising committee, pictured below (with the exception of Joanne, who is missing from the photo). I would like to extend my sincere thanks to each of them, as well as to our photographer, Can Li.

When we needed to adapt quickly due to the unfavourable weather forecast for the outdoor tours, Quan and Fonti quickly arranged for an alternative tour at the Australian Parliament House. Since this coincided with a Parliamentary sitting week, delegates had the unexpected opportunity to attend Question Time, which, from all accounts, was an enjoyable experience! Well done team in adapting so quickly!

Congratulations are in order for:

• Maggie for winning the People’s Choice Poster Prize on her work for her R package ggincerta for uncertainty visualisation with a layered grammar of graphics, and
• Jenny for receiving the Honourable Mention for the Student Poster Prize based on her Honours project on Factor Analytic Mixed Models and Plackett-Luce Models for Multi-Environmental Trials.

Well done Maggie and Jenny!

Good job also Fonti for delivering her first talk at a Biometrics conference on her heritable R package for heritability calculations for breeding trials.

We left the ANU campus in the morning for the one-hour drive to BARS. The views turned spectacular as we approached the station. It was the first time for me seeing such a massive canola field stretching into the horizon, absolutely stunning.

We were welcomed with talks from Georgie, Stuart, Susie, Aswin, John, Greg, and Allison, who introduced the research capabilities and ongoing projects at BARS. I was amazed at how modern agricultural research has become. The farm has a comprehensive sensor network, fully digitized data and research management, and even drones scanning fields to phenotype biomass.

Dr. Susie Sprague presented her work on blackleg canola disease trials, which our AAGI team is also involved in. Patrick will be developing computer vision methods to identify diseased canola plants, while I will be working on sample size calculations for QTL analysis to identify loci associated with disease resistance. Canola is one of the major crops in the ACT, and blackleg is a severe fungal disease that evolves quickly, spreads rapidly, and easily develops fungicide resistance. Susie shared the story of a local farmer struggling with blackleg who turned to BARS for help. It’s such a boost of motivation for me of how research like this can directly impact people’s lives: glad I’m choosing the correct job haha.

During the lunch break, Patrick, being much more outgoing than me, introduced himself to Susie and kindly introduced me as well. Susie explained some of the challenges in QTL analysis for this project. The blakcleg resistance trait in canola is extremely variable. Environmental factors, treatments, genetics, and measurement error together only explain a small fraction of this variability. Even plants within the same plot can show vastly different resistance levels. It would be interesting to know why of course (epigenetics, perhaps?).

We toured the farm facilities in the afternoon and saw equipment that most people never see in their lives: seed dryers, grain cleaners, and huge agricultural machinery like seeders, harvesters, and sprayers that usually only see in movies or video games. You really can’t imagine their size until you see them up close. My aim is to get on them once in these two years.

We then visited several research trials. Dr. John Kirkegaard showed us his long-term trial using old-school agronomic (no chemical, pure crop management). Faba beans are intercropped and rotated with canola, improving soil quality by fixing nitrogen. Susie took us to her canola plots, where we saw what diseased plants actually look like. Even the tiniest leaf lesion can be an entry point for the fungus. Finally, Dr. Greg Rebetzke showed us his dwarf wheat trail, where they molecular interventing the gibberellic acid signaling pathway to promote dwarf wheat phenotypes to increase yield. Those plants are only 10–20 cm tall (normally they are aorund 1 meter).

At the end of the day, we headed back home tired but happy. Others in the car asked about who we were on the drive. It was definitely easier to explain once we mentioned BDSI and Emi. Hopefully one day BDSI will return! (it was quite funny how Tony forgot Patrick’s name for the third time).

I also thoroughly enjoyed listening to the student presentations, which showcased impressive research of analytical methods to draw insight from multiomics crop data. In particular, learning about the lab’s multiomics soybean dataset left me both inspired and a little envious—I would love the opportunity to analyse such a rich source of data myself! I was also impressed with Hayato, a soon to finish PhD student, who developed the rrda R package as part of his research. Thank you again to everyone in the lab for a warm welcome and stimulating scientific exchange.

The food in Japan is incredibly delicious, and I’m amazed at how much more affordable it is compared to Australia (often about half the price)! The quality of vegetables and fruits on display in supermarkets is also outstanding. With so many tempting options, I have to admit, I’m definitely indulging a bit too much!

It’s been seven years since my last visit to Japan, and experiencing small moments like catching a glimpse of Tokyo Tower or stumbling upon a tiny roadside jinja continually reminds me of the Japanese culture. I’m looking forward to visiting again and hopefully more regularly!

The future climate conditions tend to hasten the wheat growth. You can see this in the photos below.

Thank you to Zixiong for the tour and looking forward to seeing more about this year’s data!

• the analysis of a single field trial,
• the analysis of a multi-environmental field trial,
• workflow and tools for reporting (including how we structure our projects and name things!), and
• linear mixed models.

Thank you to Fonti for showcasing about project structure and tools for project management! A happy coincidence meant we got to celebrate Fonti’s birthday with a cake in our last training session 😊

And in case if you were wondering, we did do the training! Lots of great discussions and what a load of fun learning different perspectives.

All our training materials is available here and we look forward to developing and accumulating more materials for the community!

Yidi has an impressive list of work from his PhD in developing and applying statistical methods for single cell data integration analysis. His work from his PhD include:

• Sincast - a computational framework to query scRNA-seq data (SIN) by projecting onto bulk reference atlases
• StableMate - a variable selection for multi-environmental regression
• NeighbourNet - predict gene regulatory network for individual cells in single cell RNA-seq data
• Application to human inner ear macrophage

We are looking forward to learning more about his work and working with him 😄

Fonti has a wealth of experience in R programming and developing tools that aid research for others, in addition to her experience in evolutionary ecology where she quantified the heritability of traits from her lizard breeding experiment. She’s the lead developer for the austraits R package, which won a software award. We’re looking forward to working with her and the new perspectives and ideas she will bring to the project!

Patrick, who moved from Melbourne, has developed computer vision models for automating reading of residual plots, implemented as the R package autovi. He also created the spotoroo R package, an algorithm designed to cluster satellite hot spot spatio-temporal data. His work on spotoroo earned him a software award.

The three postdocs will be part of our institute working on the “Analytics for the Australian Grains Industry” project each working on one of the following topics:

• Statistical design and modelling,
• Bioinformatics and molecular phenotyping, or
• Machine learning and image analytics.

The job application closes on Jan 19 2025 23:55 AEDT. The term of employment is up to 3 years (up to 31st December 2027).

For more details see here.