The BiGG Models knowledge base (http://bigg.ucsd.edu) is a centralized repository for high-quality genome-scale metabolic models. For the past 12 years, the website has allowed users to browse and search metabolic models. Within this update, we detail new content and features in the repository, continuing the original effort to connect each model to genome annotations and external databases as well as standardization of reactions and metabolites. We describe the addition of 31 new models that expand the portion of the phylogenetic tree covered by BiGG Models. We also describe new functionality for hosting multi-strain models, which have proven to be insightful in a variety of studies centered on comparisons of related strains. Finally, the models in the knowledge base have been benchmarked using Memote, a new community-developed validator for genome-scale models to demonstrate the improving quality and transparency of model content in BiGG Models.

Conjugation-break spacers (CBS) have been shown to enhance the mechanical properties of conjugated polymers. In particular, incorporation of CBS units into semi-random polymers has revealed high ductility and low elastic moduli, attributed to the combined influence of the CBS units and the semi-random architecture. To further elucidate the structure–property relationships in these polymers, two new families of semi-random polymers are reported here. In the first, poly­(3-hexylthiophene)-based semi-random polymers incorporating diketo­pyrrolopyrrole (DPP) units were synthesized in which CBS units with 4–10 carbons were incorporated from 10 to 40% with an equivalent content of 2-decyltetradecyl-DPP (dtdDPP) to overcome solubility limitations previously observed with 2-ethylhexyl-DPP (ehDPP). These polymers had much higher solubility and could attain higher molecular weights, formed films with high integrity, and displayed extraordinary mechanical properties, with elastic moduli as low as 5.45 MPa and fracture strains as high as 398%. In the second family, the content of ehDPP was held constant at 10%, while the CBS content was varied from 10 to 50% (with an eight-carbon spacer) to deconvolute the influence of CBS and DPP content on mechanical properties. Polymer solubility, molecular weight, and processability were not shown to improve dramatically relative to the previous generation of ehDPP polymers with matched DPP and CBS content, but the mechanical properties of this series were quite notable, with elastic moduli as low as 4.08 MPa, an increase in toughness, and fracture strains as high as 432%. The extraordinary mechanical properties exhibited by these polymers can serve as a guide in the judicious selection of monomers and backbone architectures in the future synthesis of semiconducting polymers for flexible electronic applications.

Background Independent component analysis is an unsupervised machine learning algorithm that separates a set of mixed signals into a set of statistically independent source signals. Applied to high-quality gene expression datasets, independent component analysis effectively reveals both the source signals of the transcriptome as co-regulated gene sets, and the activity levels of the underlying regulators across diverse experimental conditions. Two major variables that affect the final gene sets are the diversity of the expression profiles contained in the underlying data, and the user-defined number of independent components, or dimensionality, to compute. Availability of high-quality transcriptomic datasets has grown exponentially as high-throughput technologies have advanced; however, optimal dimensionality selection remains an open question. Methods We computed independent components across a range of dimensionalities for four gene expression datasets with varying dimensions (both in terms of number of genes and number of samples). We computed the correlation between independent components across different dimensionalities to understand how the overall structure evolves as the number of user-defined components increases. We then measured how well the resulting gene clusters reflected known regulatory mechanisms, and developed a set of metrics to assess the accuracy of the decomposition at a given dimension. Results We found that over-decomposition results in many independent components dominated by a single gene, whereas under-decomposition results in independent components that poorly capture the known regulatory structure. From these results, we developed a new method, called OptICA, for finding the optimal dimensionality that controls for both over- and under-decomposition. Specifically, OptICA selects the highest dimension that produces a low number of components that are dominated by a single gene. We show that OptICA outperforms two previously proposed methods for selecting the number of independent components across four transcriptomic databases of varying sizes. Conclusions OptICA avoids both over-decomposition and under-decomposition of transcriptomic datasets resulting in the best representation of the organism’s underlying transcriptional regulatory network.