General information on the Unit

ECTS: 3
Contact hours: 35 (15 lectures, 20 practicals)
Personal work hours: 65
Character: Compulsory
Venue: Organized by the Mediterranean Agronomic Institute of Zaragoza (CIHEAM Zaragoza)
Scheduling:
- Developed during the first academic year of the Master, at the first part of the second semester.
- The assessment of this Unit consists in a group presentation.
Requisites and permanence
Knowledge on basic statistics and R software is required. Propaedeutic on basic statistics and R software are organized for students who need them.
Learning methods
Combination of theoretical and practical classes, individual study and work.
Language
Lectures are delivered in English and the documents supplied by the lecturers are also in English. The presentation of the group work is in English. 

Presentation of the Unit and context within the syllabus

This Unit introduces infrastructure needed for field and indoor platform phenomics. Then specific experimental designs and corresponding mixed models treated in detail together with spatial and longitudinal modelling. Statistical and machine learning techniques presented for pre-processing of phenomic data. Methodologies for the identification of the genetic basis of new phenotypic traits demonstrated. Finally, phenomic traits integrated in prediction models for yield. Examples and exercises used real data from phenotyping platforms and field experiments. 

Competences

Specific competences

• SC1 Familiarize with different platforms for plant phenotyping.
• SC3 Having a good basics in indoor experimental designs.
• SC4 Assessing the relevant methods and techniques in the prediction of yield.

General competences

• GC1 Integrating scientific and technical knowledge and applying them discerningly.
• GC2 Performing scientific and/or technical information searches and processing them selectively.
• GC3 Analyzing results or strategies and elaborating conclusions which contribute to clarify the problems and to find possible solutions.

Learning outcomes

The student, at the end of the learning of this Unit:

• Knows the different platforms, sensors and carriers used for plant phenotyping.
• Be able of determining suitable experimental designs and perform analysis using mixed models.
• Be capable to correct extracted plant features for spatial and temporal trends.
• Appreciates the potential of integrating secondary plant traits in genetic models for prediction of yield.
• Have acquired practical experience in applying statistical methods through the analysis of case studies and hands-on exercises.

Contents

• Introduction to Phenotyping
• Introduction to HTP data indoor & outdoor phenotyping
• Data annotation and organization
• Phenotyping VL – Spectral data LAB
• Choosing the design for phenotyping experiments: Procedure and examples of indoor experiments designs
• Feature extraction
• Correcting for design factors and spatial modelling
• Modelling dependence on environment gradients
• Target trait prediction
• Integrating phenotyping, machine learning, crop growth, modelling and satellite imagery in plant breeding programs

Learning activities

Learning activity 1: Lectures combined with examples
ECTS: 2
Hours: 50
Percentage of contact: 30%

Learning activity 2: Tutored individual work
(a) Evaluate the spatial field design using the long data format
(b) Evaluate data with multiple measurements at multiple treatment factors
(c) A basic showcase: Prediction of visual senescence scorings from spectral data
(d) Disease detection and Quantification using spectral-temporal features
(e) Choosing the design for phenotyping experiments
(f) Correlation for spatial trends over time
(g) Time series modelling
ECTS: 0.5
Hours: 12.5
Percentage of contact: 56%

Learning activity 3: Tutored group work
Students, guided by the tutor, work in groups of 5-6 persons. They receive a set of papers on genomic prediction and each group has to select three of them. Every group must draft a synthesis document for public presentation and joint discussion.
ECTS: 1,5
Hours: 37.5
Percentage of contact: 28%

Assessment methods

Assessment system: Presentation a group work related with the interpretation of three papers on genomic prediction. The score is based both on the PowerPoint document and on the presentation and defence of the work. The score is the same for all members of the group.
Weighting: 100% of the final score of the Unit

Lecturers

Jonas ANDEREGG, ETH Zürich (Switzerland)
Scott CHAPMAN, Univ. Queensland, St Lucia (Australia)
Andy HUND, ETH Zürich (Switzerland)
Emilie MILLET, WUR (The Netherlands)
Lukas ROTH, ETH Zürich (Switzerland)
Fred Van EEUWIJK, WUR (The Netherlands)
Rick Van DE ZEDDE, WUR (The Netherlands)