Interactive graphs to help breeders take decisions

Since the origin of selective breeding, handling genetic material has always been a challenge¹. And nowadays managing germplasm diversity becomes more and more important for climate-smart breeding and to meet new consumer demands such as food security and product quality. Introducing new genetic diversity becomes crucial to find new disease and pest resistances or quality traits and abiotic stress tolerances. 

When studying hybrid phenotypic and genotypic data, graphical representations, answer a broad range of questions: Choose next season hybrids, evaluate costs, lands, manage resources... Dynamic visualization tools enable breeders to display and manage large amounts of heterogeneous data generated in a material’s lifecycle. Let's discuss some innovative ways to study germplasm, find interesting crosses, and finally put in place an organization into the breeding department to reach these objectives.

“Plant breeding must be able to meet a wide range of requirements, hence the importance of genetic diversity and the management of germplasm.” Magloire OTEYAMI²

Interactive matrix graph for hybrid breeding

Filtering parents (GCA>0) in a crossing matrix

The Hybrid Comparison matrix presents all hybrid campaign results of a trial network. Female parents in raws, males in columns and every hybrid evaluated during the campaign in the inner cells. In every cell can be read the hybrid's mean of yield results. Cell color corresponds to the specific combining ability of the hybrid's parents, on a scale of four  colors.

The matrix is interactive, we can use filters to remove some parents that have low GCA. In the animation, we have kept only males and females with SCA > 0. As a result, the breeder visualizes quickly the best parents for this program, and the possible crosses that could generate interesting hybrids in the next season !

“Most researchers use RnDExp® tools to handle their R&D and breeding processes, such as crosses, trials, pathology testing and seed quality. Every breeder of the team centralizes research data in the database and they share information depending on their user’s rights.” Gautier Semences

Interactive graphs to explore pedigrees

Selecting the most promising progenies of a hybrid

A multi-windowing tool enables to visualize heterogeneous data, such as the pedigree of a selected hybrid on the one hand, and on the other hand the performances of  this hybrid's "relatives" such as graphs, raw data and statistics. Such a tool supports breeders in finding correlations {pedigree x agronomic traits} and best new descendants to evaluate.

Top-left: All the hybrids in the trial network, with information on the genotype, general means.

Top-right: General pedigree information. In green at the center is the hybrid we are studying. 

Bottom-left: Graphical representation of the performances of all the "related hybrids":  Yield and Moisture.

Bottom-right: Raw data results from all experiments of the trial network for the selected hybrids.

For instance, such graphs have been used in a publicated study of 90 hybrids resulting from 200 starting crosses, evaluated on 5 locations. The aim was to compare yield results and calculation of GCA/SCA.³

“RnDExp® covers our experimental needs well, particularly when it comes to data analysis and labeling. [...] Moreover, as far as time efficiency, data access, accessing multiple tables at once, etc, the program is very efficient.” Jean Baptiste Guignaudeau, In charge if IT for Research operations, Unisigma

Dynamic graphs enhance breeding decision-making

RnDExp Viz® comparative analysis web app

Researchers and managers highly appreciate real time dashboards and standardized reports to pilot the department and take informed decisions. From starting crosses to product development, every breeding process benefits from actionable graphical breeding tools and geographical representations.

Image: The web application RnDExp Viz® displays data from the whole research database. This tool for data visualization and comparative analysis, included in RnDExperience® software, has been designed in an ergonomic, pedagogic and interactive way. It enables any researcher to query and report all the R&D department information, according to its user rights of course !

A transversal research information system is necessary to query all the department information in real time and get the most of digital decision-making tools. The collaborative plant breeding software enables communication between breeding, testing, laboratory and management, for the whole team to work in synergy. Such collaboration requires a flexible software configured to the specificities of every activity and their various species, material and objectives.

“Terres Inovia today disposes of an integrated data management system "from the field to the technical reference allowing time and reliability gains.” Terres Inovia annual report 2017

To go further, watch the video o our webinar :

5 strategies to get the most of your genetic variability

When studying hybrid phenotypic and genotypic data, it is challenging to find the right graphical representations, to answer a broad range of questions.

The difficulty stems from the large amount of heterogeneous data generated in a material’s lifecycle that needs to be displayed and managed.

Featuring:
Our agronomy and IT Expert Tristan Duminil, Agronomy Engineer and Business Manager.

Notes:

1 Starch grain and phytolith evidence for early ninth millennium B.P. maize from the Central Balsas River Valley, Mexico Piperno et al. 2009

2 Evaluation and Breeding of lowland rice varieties for their yield and tolerance to biotic and abiotic stresses in Benin and Togo: implication for genetic improvement of rice and food security in West Africa, Oteyami et al. 2019"

3 How interactive genealogy graphs enable to explore a hybrid's extended family and study the performances of its members for mutliple traits, all within RnDExp link