Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disease with substantial heterogeneity in its clinical presentation. This makes diagnosis and effective treatment difficult, so better tools for estimating disease progression are needed. Here, we report results from the DREAM-Phil Bowen ALS Prediction Prize4Life challenge. In this crowdsourcing competition, competitors developed algorithms for the prediction of disease progression of 1,822 ALS patients from standardized, anonymized phase 2/3 clinical trials. The two best algorithms outperformed a method designed by the challenge organizers as well as predictions by ALS clinicians. We estimate that using both winning algorithms in future trial designs could reduce the required number of patients by at least 20%. The DREAM-Phil Bowen ALS Prediction Prize4Life challenge also identified several potential nonstandard predictors of disease progression including uric acid, creatinine and surprisingly, blood pressure, shedding light on ALS pathobiology. This analysis reveals the potential of a crowdsourcing competition that uses clinical trial data for accelerating ALS research and development

In the sentence beginning 'The meQTL are enriched for functionally relevant characteristics', please provide a de inition for Hi-C.2. In the author list, there are two authors with the name Panos Deloukas with different af iliations. Please con irm whether this is the same individual and whether the two authors can be merged. 3.Please check your article carefully, coordinate with any co-authors and enter all inal edits clearly in the eproof, remembering to save frequently. Once corrections are submitted, we cannot routinely make further changes to the article. 4. Note that the eproof should be amended in only one browser window at any one time; otherwise changes will be overwritten. 5. Author surnames have been highlighted. Please check these carefully and adjust if the irst name or surname is marked up incorrectly. Note that changes here will affect indexing of your article in public repositories such as PubMed. Also, carefully check the spelling and numbering of all author names and af iliations, and the corresponding email address(es).6. You cannot alter accepted Supplementary Information iles except for critical changes to scienti ic content. If you do resupply any iles, please also provide a brief (but complete) list of changes. If these are not considered scienti ic changes, any altered Supplementary iles will not be used, only the originally accepted version will be published. If applicable, please ensure that any accession codes and datasets whoseDOIs or other identi iers are mentioned in the paper are scheduled for public release as soon as possible, we recommend within a few days of submitting your proof, and update the database record with publication details from this article once available. 8. Your paper has been copy edited. Please review every sentence to ensure that it conveys your intended meaning; if changes are required, please provide further clari ication rather than reverting to the original text. Please note that formatting (including hyphenation, Latin words,and any reference citations that might be mistaken for exponents) has been made consistent with our house style. 9. Please con irm or correct the city name inserted in af iliations 5, 10, 12, 13, 21, 22, 39 and 44. 10. In the sentence beginning 'Our meQTL replicate in data generated by the Illumina', please provide a de inition for EPIC and con irm whether the de inition provided for MeDIP-seq is correct. 11. In the legend for Fig. 2g, please provide a de inition for SAT. 12. In the sentence beginning 'We used summary data-based Mendelian', please con irm whether it is correct to say '0.05 ÷ n tests'. 13. In the sentence beginning 'We then tested the 4,811 eQTL genes', please provide a de inition for coloc PP4. 14. Please note, we reserve ‘signi icant' and its derivatives for statistical signi icance. This has been reworded where this was not the intended meaning (for example to important, notable, ...

A major goal in systems biology is a comprehensive description of the entirety of all complex interactions between different types of biomolecules—also referred to as the interactome—and how these interactions give rise to higher, cellular and organism level functions or diseases. Numerous efforts have been undertaken to define such interactomes experimentally, for example yeast-two-hybrid based protein-protein interaction networks or ChIP-seq based protein-DNA interactions for individual proteins. To complement these direct measurements, genome-scale quantitative multi-omics data (transcriptomics, proteomics, metabolomics, etc.) enable researchers to predict novel functional interactions between molecular species. Moreover, these data allow to distinguish relevant functional from non-functional interactions in specific biological contexts. However, integration of multi-omics data is not straight forward due to their heterogeneity. Numerous methods for the inference of interaction networks from homogeneous functional data exist, but with the advent of large-scale paired multi-omics data a new class of methods for inferring comprehensive networks across different molecular species began to emerge. Here we review state-of-the-art techniques for inferring the topology of interaction networks from functional multi-omics data, encompassing graphical models with multiple node types and quantitative-trait-loci (QTL) based approaches. In addition, we will discuss Bayesian aspects of network inference, which allow for leveraging already established biological information such as known protein-protein or protein-DNA interactions, to guide the inference process.

Highlights d The glucocorticoid receptor (GR) shows rhythmic binding to chromatin in mouse liver d The majority of diurnally oscillating transcripts are directly regulated by GR d Hepatic GR controls circulating glucose and triglycerides during feeding and fasting d High-fat diet (HFD) enhances GR-STAT5 co-occupancy of metabolic enhancers at night