Machine Learning and Computational Biology Lab

The "grand finale" of our Innovative Training Network on "Machine Learning Frontiers in Precision Medicine" is taking place in person at the Max Planck Institute of Psychiatry in Munich on October 18 and 19, 2022.

Anyone can register for free, and lunch+coffee breaks are free as well!

https://mlfpm.eu/events/closing-conference/

On September 23, 2022, Max Horn defended his excellent PhD thesis entitled "Representation Learning for Dimensionality Reduction, Irregularly-​Sampled Sequences and Graphs".

Well done, Max! We wish you all the best for your next career steps.

AI offers a faster way to predict antibiotic resistance A study under co-leadership of the ETH Zurich has shown that computer algorithms can determine antimicrobial resistance of bacteria faster than previous methods. This could help treat serious infections more efficiently in the future.

On October 18, Christian Bock defended his excellent PhD thesis entitled "Motifs and Manifolds Statistical and Topological Machine Learning for Characterising and Classifying Biomedical Time Series".

Very well done, Christian! All the best for your future.

Our series of " in Medicine" summer schools features its 7th edition on September 20-22 this month, live and free on YouTube, with lots of exciting talks. You are welcome to tune in!

https://t.co/y3WpCOWwcy

3rd MLFPM Summer School General information Date: September 20-22, 2021 Place: ONLINE. The link to the livestream will be published here. Registration: No registration needed Talks on YouTube We will have a livestream on…

Open postdoctoral position: "Machine learning on structured data and its applications in the life sciences"

Application deadline: July 27, 2021.

Job Opportunities Job opportunities at the Machine Learning and Computational Biology Lab

New publication: RNA atlas of human bacterial pathogens uncovers stress dynamics linked to infection

In a collaboration with the Universities of Umeå and Helsinki, and Karolinska Institute, Matteo and Karsten investigated stress responses in many human bacterial pathogens. The work has now been published in Nature Communications.

RNA atlas of human bacterial pathogens uncovers stress dynamics linked to infection Bacterial stress responses are potential targets for new antimicrobials. Here, Avican et al. present global transcriptomes for 32 bacterial pathogens grown under 11 stress conditions, and identify common and unique regulatory responses, as well as processes participating in different stress response...

New publication: Early Prediction of Sepsis in the ICU Using Machine Learning: A Systematic Review

Michael, Bastian, Max, Catherine, and Karsten systematically reviewed and evaluated studies employing machine learning for the prediction of sepsis in the ICU.

www.frontiersin.org

Congratulations, Matteo Togninalli, on winning an ETH medal for your outstanding PhD thesis!

https://www.research-collection.ethz.ch/handle/20.500.11850/448681

New publication: Enhancing statistical power in temporal biomarker discovery through representative shapelet mining

Temporal biomarker discovery in longitudinal data is based on detecting reoccurring trajectories, the so-called shapelets. The search for shapelets requires considering all subsequences in the data. The redundancy and overlap of the detected shapelets results in an a priori unbounded number of highly similar and structurally meaningless shapelets. As a consequence, current temporal biomarker discovery methods are impractical and underpowered. In this paper, we present a novel method for temporal biomarker discovery: Statistically Significant Submodular Subset Shapelet Mining (S5M) that retrieves short subsequences that are (i) occurring in the data, (ii) are statistically significantly associated with the phenotype and (iii) are of manageable quantity while maximizing structural diversity. Structural diversity is achieved by pruning non-representative shapelets via submodular optimization. This increases the statistical power and utility of S5M compared to state-of-the-art approaches on simulated and real-world datasets. For patients admitted to the intensive care unit (ICU) showing signs of severe organ failure, we find temporal patterns in the sequential organ failure assessment score that are associated with in-ICU mortality.

This work was conducted by Thomas, Christian, Michael, Bastian and Karsten.

Enhancing statistical power in temporal biomarker discovery through representative shapelet mining AbstractMotivation. Temporal biomarker discovery in longitudinal data is based on detecting reoccurring trajectories, the so-called shapelets. The search for sh

Big news for MLCB: Karsten will join the leadership team of the Department of Biosystems Science and Engineering (D-BSSE) at ETH Zurich as Deputy Department Chair, together with Prof. Daniel Müller as Department Chair and Prof. Petra Dittrich as Director of Studies, effective early 2021.

https://bsse.ethz.ch

Excited by the current boom in learning on graphs, we have reviewed the foundations and trends in graph kernel research. We hope that it will be a reference and starting point for lots of future work in this domain!

A massive (151-page) effort by Karsten, Elisabetta, Felipe, Leslie and Bastian. Congratulations!

https://arxiv.org/abs/2011.03854