Kernel Methods, Pattern Analysis and Computational Metabolomics (KEPACO)
The KEPACO group develops machine learning methods, models and tools for data science, in particular computational metabolomics. The methodological backbone of the group is formed by kernel methods and regularized learning. The group particularly focusses in learning with multiple and structured targets, multiple views and ensembles. Machine learning applications of interest include metabolite identification, metabolic network reconstruction and pathway analysis, chemogenomics as well as biomarker discovery.
See overview of KEPACO research (in PDF)
- KEPACoffee, regular group gathering
- February 23, 2017: Our Metabold team pitched in the Helsinki Challenge pitch night. See our pitch in Youtube here, our part starts at 1:26:10
- January 16, 2017: We got to the semi-final of Helsinki Challenge with our joint team with University of Eastern Finland, lead by Dr. Kati Hanhineva. 20 ouf of 110 teams were selected for the semi-final.
- January 9, 2017: Mohamed Jabri begins as an intern in KEPACO. Welcome Mohamed!
- November 17, 2016: Sahely Bhadra presents the paper "Multi-view kernel completion" in ACML 2016 conference, Hamilto, NZ
- October 19-21, 2016: Sandor Szedmak will present the paper "Soft Kernel Target Alignment for Two-stage
Multiple Kernel Learning" at the 19th International Conference on Discovery Science (DS 2016) at Bari, Italy
- September 3-4, 2016: KEPACO Co-organize the 10th International Workshop on Machine Learning in Systems Biology (MLSB 2016), in conjunction with ECCB 2016 conference in The Hague, The Netherlands. CALL FOR PAPERS DEADLINE June 10, 2016. Proceedings track papers published in BMC Bioinformatics.
- July 8-12, 2016: Celine Brouard has an oral presentation on fast metabolite identification through input-output kernel regression at ISMB 2016, Orlando, USA
- June 1, 2016: Heli Julkunen starts as a summer intern. Welcome Heli!
- May 19-20, 2016: Anna Cichonska and Viivi Uurtio co-organize the Bioinformatics Research and Education Workshop (BREW) in Helsinki
- May 9-11, 2016: Markus Heinonen present his paper on non-stationary Gaussian process inference at AISTATS 2016, Cadiz, Spain
- May 6, 2016: KEPACO did well in the CASMI 2016 contest, Team Brouard (KEPACO/Jena) winning one category and Team Duehrkop (Jena/KEPACO) coming second.
- May 2-June 19, 2016: Juho makes a research visit to University College London
- December 18, 2015: Academy of Finland granted funding for the project "Data-Driven Decision Support for Digital Health - D4Health" as part of the ICT2023 programme. The project is implemented by consortium with 5 research groups from Aalto, FIMM and HIIT.
- December 12, 2015: Celine Brouard and Hongyu Su will present posters in the NIPS MLCB/MLSB-2015 workshop
- November 29 -- December 4, 2015: KEPACO co-organized Dagstuhl workshop Computational Metabolomics with ca. 30 international top experts
- November 4, 2015: Juho Rousu appeared in YLE Popular science program Prisma Studio (In Finnish only!), explaing how metabolites are identified using machine learning in the CSI:FingerID search engine. See the clip here! (.mp4, 50MB
- Older news
- Juho Rousu, Associate professor, group leader
- Sandor Szedmak, PhD, senior research scientist
- Celine Brouard, PhD, post-doctoral researcher
- Huibin Shen, PhD student
- Anna Cichonska, PhD student (HIIT-FIMM)
- Viivi Uurtio, PhD student
- Eric Bach, PhD student
- Parisa Mapar, MSc student
- Linh Nguyen, MSc Honours student
- Heli Julkunen, research assistant
- Mohamed Jabri, research assistant
The KEPACO group is located at the Department of Computer
Science at the School of
Science of Aalto University. We
also belong to the Helsinki Institute
for Information Technology.
Contact information and how
to get to CS department in Aalto University Otaniemi Campus
- FCHealth - Foundations of Computational Health, HIIT research programme
- D4Health - Data-Driven Decision Support for Digital Health. Academy of Finland grant 2016-2017.
- LiF - Living Factories. Finnish Funding Agency for Innovation (1. phase 2014-2016)
- MIDAS - Metabolite Identification through Algorithms and Statistical Learning. Academy of Finland grant 2013-2017
Check out the CSI:FingerID server for metabolite identification from MS/MS data, running the methods we developed with Sebastian Boecker's group in Friedrich-Schiller-Universitat Jena.
You may also check the clip from YLE Popular science program Prisma Studio (In Finnish only!), showcasing how metabolites are identified using machine learning in the CSI:FingerID search engine. See the clip here! (.mp4, 50MB
Please find our software page here and our GitHub page at github.com/aalto-ics-kepaco.
Selected and recent publications
- Sahely Bhadra, Samuel Kaski, Juho Rousu. Multi-view kernel completion. ACML 2016 journal track. Machine Learning, 2016, to appear
- Huibin Shen, Sandor Szedmak, Celine Brouard, Juho Rousu. Soft Kernel Target Alignment for Two-Stage Multiple Kernel Learning. Proc. 19th International Conference on Discovery Science (DS 2016), 2016, to appear
- Jana Kludas, Mikko Arvas, Sandra Castillo, Tiina Pakula, Merja Oja, Celine Brouard, Jussi Jantti, Merja Penttila, Juho Rousu. Machine learning of protein interactions in fungal secretory pathways. PLOS One, 2016, to appear
- Celine Brouard, Huibin Shen, Kai Duehrkop, Florence D'Alche-buc, Sebastian Boecker and Juho Rousu. Fast metabolite identification with Input Output Kernel Regression. Proc. Intelligent Systems for Molecular Biology, ISMB 2016, Bioinformatics, 2016, to appear
- Markus Heinonen, Henrik Mannerstrom, Juho Rousu, Samuel Kaski, Harri Lahdesmaki. Non-Stationary
Gaussian Process Regression with Hamiltonian Monte Carlo. The 19th International Conference on
Articial Intelligence and Statistics, AISTATS 2016, Journal of Machine Learning Research Workshop
and Conference Proceedings, 2016, to appear
- Honeyborne I, McHugh TD, Kuittinen I, Cichonska A, Evangelopoulos D, Ronacher K, van Helden PD, Gillespie SH, Fernandez-Reyes D, Walzl G, Rousu J. Profiling persistent tubercule bacilli from patient sputa during therapy predicts early drug efficacy. BMC medicine. 2016 Apr 7;14(1):1.
- Anssi Rantasalo, Elena Czeizler, Riitta Virtanen, Juho Rousu, Harri Lahdesmaki, Merja Penttila, Jussi Jantti, Dominik Mojzita. Synthetic Transcription Amplifier System for Orthogonal Control of Gene Expression in Saccharomyces cerevisiae. PLOS One, 2016, doi:10.1371/journal.pone.0148320
- Viivi Uurtio, Malin Bomberg, Kristian Nybo, Merja Itavaara, Juho Rousu.
Canonical Correlation Methods for Exploring Microbe-Environment Interactions in Deep Subsurface. Discovery Science, 2015. Lecture Notes in Computer Science, vol 9356, pp. 299--307
- Anna Cichonska, Juho Rousu, Pekka Marttinen, Antti J Kangas, Pasi Soininen, Terho Lehtimaki, Olli T Raitakari, Marjo-Riitta Jarvelin, Veikko Salomaa, Mika Ala-Korpela, Samuli Ripatti, Matti Pirinen. metaCCA: Summary statistics-based multivariate meta-analysis of genome-wide association studies using canonical correlation analysis. Bioinformatics, Advance Access February 2016.
- Kai Duehrkop , Huibin Shen, Marvin Meusel, Juho Rousu, and Sebastian Boecker. Searching molecular structure databases with tandem mass spectra using CSI:FingerID. Proceedings of the National Academy of Sciences, vol. 112, 41 (2015) pp. 12580-12585
- Anna Cichonska, Juho Rousu, Tero Aittokallio}. Identification of Drug Candidates and Repurposing Opportunities Through Compound-Target Interaction Networks. Expert Opinion On Drug Discovery, Vol. 10, Iss. 12, 2015
- Jenni Hultman, Riitta Rahkila, Javeria Ali, Juho Rousu and K. Johanna Bjorkroth: Meat processing plant microbiome and contamination patterns of cold-tolerant bacteria causing food safety and spoilage risks in the manufacture of vacuum-packaged, cooked sausages.. Applied Environmental Microbiology 81, 20 (2015), pp. 7088-7097
- Hongyu Su, Juho Rousu. Multilabel classification through random graph ensembles. Machine Learning, Volume 99, Issue 2 (2015), pp 231-256 http://dx.doi.org/10.1007/s10994-014-5465-9.
- Hongyu Su. Multilabel Classification through Structured Output Learning - Methods and Applications. PhD thesis, Aalto university, 2015
- Mario Marchand, Hongyu Su, Emilie Morvant, Juho Rousu, John Shawe-Taylor. Multilabel Structured Output Learning with Random Spanning Trees of Max-Margin Markov Networks. To appear at Neural Information Processing Systems (NIPS), 2014
- Huibin Shen, Kai Duehrkop, Sebastian Boecker, Juho Rousu. Metabolite Identification through Multiple Kernel Learning on Fragmentation Trees. Bioinformatics, 2014, accepted for presentation at ISMB-2014
- Hongyu Su, Aristides Gionis, Juho Rousu. Structured Prediction of Network Response. 31th International Conference on Machine Learning, ICML 2014, to appear.
- Esa Pitkanen, Paula Jouhten, Jian Hou, Muhammad Fahad Syed, Peter Blomberg, Jana Kludas, Merja Oja, Liisa Holm, Merja Penttila, Juho Rousu, Mikko Arvas. Comparative genome-scale reconstruction of gapless metabolic networks for present and ancestral species. PLoS computational biology 10, 2 (2014), e1003465
- Juho Rousu, Daniel D. Agranoff, Olugbemiro Sodeinde, John Shawe-Taylor, Delmiro Fernandez-Reyes. Biomarker Discovery by Sparse Canonical Correlation Analysis of Complex Clinical Phenotypes of Tuberculosis and Malaria. PLOS Computational Biology 9 (4), 2013, e1003018
Visitors to the group
- 2016: Prof. Sebastian Boecker, Dr. Tim White, Marcus Ludwig, Kai Duehrkop, Friedrich-Schiller University Jena
- 2015: Prof. Giorgio Valentini, Universita Degli Studi di Milan
- 2013: Prof. Sebastian Boecker, Friedrich-Schiller University Jena
- Dr. Ari Rantanen, PhD 2006, currently at Sanoma Corp.
- Dr. Esa Pitkänen, PhD 2010, currently at Aaltonen Lab, University of Helsinki
- Dr. Markus Heinonen, PhD 2013, currently at CSB group, Aalto University
- Dr. Jefrey Lijffijt, PhD 2013, moved to University of Bristol
- Dr. Jana Kludas, left for DSSG Berlin
- Dr. Hongyu Su, PhD 2015, left for Nordea Bank
- Dr. Elena Czeizler, research fellow, 2013-16, left for Citrus solutions
- Dr. Sahely Bhadra, post-doc, 2014-2016, left for Northeastern University, Boston, USA
- Yvonne Herrmann, MSc 2012
- Fitsum Tamene, MSc 2013
- Jian Hou, MSc 2014
- Iitu Kuittinen, MSc 2015
- Nicole Althermeler, MSc 2016
- Jinmin Lei, MSc 2016
- Maja Ilievska, MSc 2016
- Jane Brodie, 2014-15
- Clemens Westrup, 2013-15
- Carlos Maycas Nadal, BSc 2014