Real-World Molecular Out-Of-Distribution: Specification and Investigation.

This study presents a rigorous framework for investigating molecular out-of-distribution (MOOD) generalization in drug discovery. The concept of MOOD is first clarified through a problem specification that demonstrates how the covariate shifts encountered during real-world deployment can be characterized by the distribution of sample distances to the training set. We find that these shifts can cause performance to drop by up to 60% and uncertainty calibration by up to 40%. This leads us to propose a splitting protocol that aims to close the gap between the deployment and testing. Then, using this protocol, a thorough investigation is conducted to assess the impact of model design, model selection, and data set characteristics on MOOD performance and uncertainty calibration. We find that appropriate representations and algorithms with built-in uncertainty estimation are crucial to improving performance and uncertainty calibration. This study sets itself apart by its exhaustiveness and opens an exciting avenue to benchmark meaningful algorithmic progress in molecular scoring.

Fission yeast kinesin-8 controls chromosome congression independently of oscillations.

In higher eukaryotes, efficient chromosome congression relies, among other players, on the activity of chromokinesins. Here, we provide a quantitative analysis of kinetochore oscillations and positioning in Schizosaccharomyces pombe, a model organism lacking chromokinesins. In wild-type cells, chromosomes align during prophase and, while oscillating, maintain this alignment throughout metaphase. Chromosome oscillations are dispensable both for kinetochore congression and stable kinetochore alignment during metaphase. In higher eukaryotes, kinesin-8 family members control chromosome congression by regulating their oscillations. By contrast, here, we demonstrate that fission yeast kinesin-8 controls chromosome congression by an alternative mechanism. We propose that kinesin-8 aligns chromosomes by controlling pulling forces in a length-dependent manner. A coarse-grained model of chromosome segregation implemented with a length-dependent process that controls the force at kinetochores is necessary and sufficient to mimic kinetochore alignment, and prevents the appearance of lagging chromosomes. Taken together, these data illustrate how the local action of a motor protein at kinetochores provides spatial cues within the spindle to align chromosomes and to prevent aneuploidy.

MAARS: a novel high-content acquisition software for the analysis of mitotic defects in fission yeast.

Faithful segregation of chromosomes during cell division relies on multiple processes such as chromosome attachment and correct spindle positioning. Yet mitotic progression is defined by multiple parameters, which need to be quantitatively evaluated. To study the spatiotemporal control of mitotic progression, we developed a high-content analysis (HCA) approach that combines automated fluorescence microscopy with real-time quantitative image analysis and allows the unbiased acquisition of multiparametric data at the single-cell level for hundreds of cells simultaneously. The Mitotic Analysis and Recording System (MAARS) provides automatic and quantitative single-cell analysis of mitotic progression on an open-source platform. It can be used to analyze specific characteristics such as cell shape, cell size, metaphase/anaphase delays, and mitotic abnormalities including spindle mispositioning, spindle elongation defects, and chromosome segregation defects. Using this HCA approach, we were able to visualize rare and unexpected events of error correction during anaphase in wild-type or mutant cells. Our study illustrates that such an expert system of mitotic progression is able to highlight the complexity of the mechanisms required to prevent chromosome loss during cell division.