Pre-conference courses

Pre-conference courses (Monday June 6)


Morning courses (9:00-12:45h)

agefederico ANOVA-SCA (ASCA) and Multilevel-SCA (MSCA).

COURSE:
Analytical chemistry, especially in the context of multidisciplinary applications, has been witnessing an increase in problems where multivariate profiles are collected as a result of designed experiments. Metabolomics and the other –omics disciplines are the domain of election, but problems of such kind are emerging more and more often in food and environmental chemistry or ecotoxicology, just to cite a few. To deal with these issues, specific multivariate tools have been proposed: ANOVA-Simultaneous Component Analysis (ASCA) and Multilevel Simultaneous Component Analysis (MSCA) are generalizations of (M)ANOVA fixed-effect models for high-dimensional data, which are being increasingly used in different areas of chemometrics and -omics.

The course will provide an introduction to the theory and application of these tools and it will be articulated along the following outline:

  1. Introduction to the theory of linear models.
    Factorial designs: fixed vs random effects; crossed, nested, replicated measurements and mixed designs. Contrasts, calculation of the effects and testing (45 minutes).
  2. Generalization to high-dimensional data from crossed –designs.
    ANOVA-SCA (ASCA): model and calculation of the effect matrices. Validation of the significance of the effects. Interpretation of the model. Graphical tools (60 minutes).
  3. Generalization to high-dimensional data from nested –designs.
    Multilevel-SCA (MSCA): model and calculation of the effect matrices. Validation of the significance of the effects. Interpretation of the model. Graphical tools. Supervised modeling of the within group variation (Multilevel-PLSDA) (60 minutes).

All ideas will be illustrated with real-life examples from chemistry and metabolomics.


Lecturers:
Age Smilde

BIO:
Age Smilde is professor of Biosystems Data Analysis at the University of Amsterdam and is also part-time professor at Copenhagen University. The ASCA method was developed in his group and he was closely involved in the development of MSCA.

(University of Amsterdam),
Federico Marini

BIO:
Federico Marini is researcher and professor of chemometrics at the University of Rome “La Sapienza”. His research interests involve, among others, the design and application of linear and non-linear classification methods, especially to food science and metabolomics.

(Università degli studi di Roma ‘La Sapienza’, Roma).

Room: Nicolau d’Olwer
ferrer Introduction to Multivariate Statistical Process Control (MSPC)

COURSE:
Goals

  • To introduce MSPC as a powerful tool for multivariate process monitoring and fault detection & diagnosis, and process improvement. 
  • To emphasize the strategic role of MSPC in Process Analytical Technology (PAT)

Target audience
The workshop is addressed to anyone interested in discovering a successful chemometrics tool for process understanding and improvement
Outline

  • What is MSPC?
  • MSPC and PAT
  • Latent Variables-based MSPC: application to continuous & batch processes

Other information
Participants are encouraged to bring their own laptops with the demo version of SIMCA software (available from http://www.umetrics.com/simca) and/or the demo version of PROMV software (available from http://www.prosensus.ca).

Note that academic users can get free licenses of ProMV, on request.


Lecturer:
Alberto Ferrer

BIO:
Dr. Alberto Ferrer is a Professor of Statistics at the Department of Applied Statistics, Operation Research and Quality, and Head of the Multivariate Statistical Engineering Research Group (mseg.webs.upv.es/index.html) at Technical University of Valencia (Spain). His main interest focuses on statistical techniques for quality and productivity improvement, especially those related to multivariate statistical methods for both continuous and batch processes. He is active in industrial teaching and consultancy activities on Big Data, Six Sigma, Process Analytical Technology (PAT) and Process Chemometrics.

(Universitat Politècnica de València, UPV).

Room: Fontseré
joaquimannadejuan Multivariate Curve Resolution (MCR).
COURSE:
The course will be a combination of theoretical concepts and hands-on work around the topic of Multivariate Curve Resolution, particularly the description and the use of the MCR-Alternating Least Squares (MCR-ALS) algorithm. There will be explanations on how to work with single data sets and multiset structures (formed by several data tables together). Recent variants of MCR incorporating hard-modeling information (e.g., kinetic laws,..) and calibration tasks (correlation constraint) will also be described. Practical examples will include analytical data (chromatography, processes,…)  and hyperspectral images. Theory and hands-on work will be combined during the course.
Program
  • MCR. Basic concept. MCR-ALS algorithm. Constraints.
  • Single data analysis.
  • Multiset analysis.
Documentation, software and data sets will be provided.
Requirements: It is recommended that attendants have some basic background in multivariate data analysis. Since practical work will be done, bringing a laptop is necessary to follow adequately the course. MATLAB is recommended, but not compulsory. The GUI interface provided can run under MATLAB environment or as a stand-alone in the compiled form.

Lecturers:
Joaquim Jaumot
BIO:
Joaquim Jaumot, senior researcher at the Spanish Council of Research (CSIC) in Barcelona. He has worked as a professor at the Universitat de Barcelona and has a long experience in the MCR topic, particularly in the analysis of biochemical processes and –omic studies. He is the designer and main developer of the GUIs linked to the MCR algorithm.
 (IDAEA-CSIC, Barcelona)
Anna de Juan
BIO:
Anna de Juan, associate professor at the Universitat de Barcelona. She has more than 20 years of experience developing and applying the MCR technique in very diverse areas. She has experience in teaching MCR in undergraduate and graduate programs at different international institutions and also in conferences and private companies.
 (Universitat de Barcelona).

Room: Puig i Cadafalch
beata Data preprocessing.
COURSE:
Data analysis is a multistage process. All its elements are equally important and decide about final results and conclusions. There are strict rules concerning sampling (in a case of natural products), experimental design, data modeling, and results validation. There are not, however, any rules concerning data pre-processing. The preprocessing step usually involves preprocessing of individual signals (signals enhancement via signals de-nosing and background elimination), as well as preprocessing of the signal set (signal transformations, normalization, alignment, etc.). All steps of data pre-processing are data dependent and they should be based on data characteristics. An influence of different data noise types, different normalization methods, different transformations, etc., on a final identification of the significant features will be discussed for the real and simulated data sets. Certain diagnostics tools will be introduced to evaluate data properties and simple rules will be proposed concerning an order of the pre-processing steps. Moreover, a comparison of data processing based on peak table and on entire signals (fingerprints) will be presented.

Course duration: 4 hours.

Lecturer:
Beata Walczak
(University of Silesia, Katowice).
BIO:
From the early 90’s Prof. Beata Walczak has been involved in chemometrics and her main scientific interests are in all aspects of data exploration and modelling (dealing with missing and censored data, dealing with outliers, data representativity, enhancement of instrumental signals, signal warping, data compression, linear and non-linear projections, development of modeling approaches, feature selection techniques etc.). She authored and co-authored ca. 160 scientific papers and delivered many invited lectures at the numerous international chemistry meetings. She acted as an co-editor of four-volume Comprehensive Chemometrics, Elsevier, Amsterdam, 2009. Currently she acts as Editor of the journal Chemometrics and Intelligent Laboratory Systems and of ‘Data Handling in Chemistry and Technology’ (the Elsevier book series), and also as a member of the editorial boards of Talanta, Analytical Letters, J. Chemometrics, Acta Chromatographica and Advances in Analytical Chemistry.


Room: Pi i Sunyer
Afternoon courses (14:00–17:45h)
riccardo Introduction to Design of Experiments, DoE
COURSE:
The main goal of this introductory course is to show the participants that Experimental Design is based on the knowledge of the problem much more than on mathematical or statistical formulas.
It will be shown that the key of the success of an Experimental Design is the chemical knowledge, and that the writing of the experimental matrices and the statistical computations can be very easily performed by hand. Under these premises, it will be understood that the specific software (very often not even required) is just a tool, not the center of the procedure.
The main topics treated will be:
  • Why optimizing OVAT (One Variable At a Time) is wrong
  • Full Factorial Designs
  • Plackett-Burman Designs
  • Central Composite Designs

Lecturer:
Riccardo Leardi
BIO:
Riccardo Leardi was born in Novi Ligure (Italy) on October 17, 1959.
In 1983 he graduated cum laude in Pharmaceutical Chemistry and Technology at the Faculty of Pharmacy of the University of Genova.

His actual position is Associate Professor at the Department of Pharmacy of the School of Medical and Pharmaceutical Sciences of the University of Genoa. In 2013 he got the qualification for full professor in Analytical Chemistry.
Since 1985 he has been working in the section of Analytic Chemistry of the Department of Pharmaceutical and Food Chemistry and Technology of the Faculty of Pharmacy of the University of Genova, and his research field is Chemometrics.

His interests are mainly devoted to problems of classification and regression (applied especially to food, environmental and clinical data), experimental design, process optimization, multivariate process monitoring and multivariate quality control.

He developed a general purpose chemometrical software, freely downloadable from the site http://gruppochemiometria.it/gruppo-lavoro-r-in-chemiometria.html

He is author of more than 120 papers and more than 120 communications in national and international meetings.

Since November 2002 he started his activity of chemometric consultancy.
(Università degli Studi di Genova, Genova).

Room: Pi i Sunyer
preys Multiblock analysis

COURSE:
This course addresses scientists who want to investigate information from combining different kinds of datasets (ex.: NIR, MIR, Raman, process parameters, sensory, HPLC, GC, …).

It will cover the following topics:

  • Introduction to multiblock analysis principles
  • Review of the main multiblock methods    
  • Application case using Matlab® environment (hands-on work).

It is therefore recommended to bring its own laptop with a valid Matlab® (Mathworks) license (no specific toolbox required).


Lecturer:
S. Preys
(Ondalys, France)
BIO:
Dr.Sebastien Preys is a chemometrician, project manager at Ondalys company in Montpellier, France.  He received his Master's degree in 1996 from Ecole d'Ingénieur Oniris, Nantes, France, and his Ph.D in Analytical Chemistry and Chemometrics from Montpellier University, France in 2006. He has been working in many different applications for the food, chemical and pharmaceutical industries and research and technical centers by providing consulting and training


Room: Puig i Cadafalch
federico Classification methods

COURSE:
Problems involving the classification of samples, i.e., the prediction of a qualitative response, usually associated to the object belonging to group of similar individuals with specified characterstics, is ubiquituous in analytical chemistry and chemometrics. Examples include the authentication of foodstuff, the evaluation of the potential toxicity of a chemical substance, the prognosis or diagnosis of a disease or, in the framework of the –omic disciplines, the identification of predictive biomarkers.
The aim of the course is to provide the attendees with a general introduction to the main classification methods through alternating theory and worked examples in Matlab.
The general outline of the course will be:

  1. Introduction to classification methods.
    What is classification? Classification vs clustering. Discriminant classification vs class-modeling; parametric vs non parametric methods. Classification error and loss function
  2. Discriminant classification methods.
    Linear and quadratic discriminant analysis and their limitation. Partial least squares discriminant analysis (PLS-DA): model building and interpretation of the results.
  3. Class-modeling
    The idea behind class-modeling. The SIMCA methods in its different implementations. Coomans plot
  4. Validation of classification methods
    The need of validating predictive models. Test set vs cross-validation. Double CV and bootstrap.

Lecturer:
Federico Marini

BIO:
Federico Marini is researcher and professor of chemometrics at the University of Rome “La Sapienza”. His research interests involve, among others, the design and application of linear and non-linear classification methods, especially to food science and metabolomics.

(Università degli Studi di Roma ‘La Sapienza’, Roma).

Room Nicolau d’Olwer
barry Multivariate Image Analysis.

COURSE:
Intro to Multivariate Image Analysis is designed to give the student practical experience. The course starts with a brief review of principal components analysis (PCA) and partial least squares (PLS) regression and how they are used in image analysis. Tools for exploring multivariate images and performing particle analysis are demonstrated, along with linking of scores in the image plane and in score-score plots. Classification of areas on images is discussed. Additional topics to be covered included multivariate image regression, and preprocessing to capture textural information. Methods to mitigate the effects of background interference, e.g. clutter, will also be discussed. The course includes hands-on computer time for participants to work example problems using PLS_ToolboxMIA_Toolbox, and MATLAB.

Lecturer:
Barry Wise

BIO:
Dr. Barry M. Wise, PLS_Toolbox creator and President and co-founder of Eigenvector Research, holds a doctorate in Chemical Engineering and has experience in a wide variety of applications spanning chemical process monitoring, modeling and analytical instrumental development. He has extensive teaching experience, having presented over 100 chemometrics courses and has co-authored over 50 peer reviewed articles, book chapters and patents. Dr. Wise is the winner of the 2001 EAS Award for Achievements in Chemometrics. He has organized and chaired numerous conferences including the First International Chemometrics InterNet Conference InCINC ’94, 1995 Gordon Research Conference on Statistics in Chemistry, Three-way Methods in Chemistry and Psychology TRICAP ’97, and Chemometrics in Analytical Chemistry, CAC-2002. Barry is also very active in the alpine ski racing community and is winner of the Pacific Northwest Ski Association’s John Genoud Award for service as a race official in 2012 and the Mission Ridge Ski Team Hampton award for volunteer service in 2014.

(Eigenvector Research, U.S.A.)

Room: Fonseré
News related to the program and structure of the courses will be published soon
logoGPBlanco

Tel. +34 932.388.777
This email address is being protected from spambots. You need JavaScript enabled to view it.

Maria Cubi, 4 - Pral.
08006 Barcelona (SPAIN)

Website create by
LogoNegro