Identifying outbreaks in sewer networks: An adaptive sampling scheme under network's uncertainty.

Motivated by the implementation of a SARS-Cov-2 sewer surveillance system in Chile during the COVID-19 pandemic, we propose a set of mathematical and algorithmic tools that aim to identify the location of an outbreak under uncertainty in the network structure. Given an upper bound on the number of samples we can take on any given day, our framework allows us to detect an unknown infected node by adaptively sampling different network nodes on different days. Crucially, despite the uncertainty of the network, the method allows univocal detection of the infected node, albeit at an extra cost in time. This framework relies on a specific and well-chosen strategy that defines new nodes to test sequentially, with a heuristic that balances the granularity of the information obtained from the samples. We extensively tested our model in real and synthetic networks, showing that the uncertainty of the underlying graph only incurs a limited increase in the number of iterations, indicating that the methodology is applicable in practice.

LoRA-TV: read depth profile-based clustering of tumor cells in single-cell sequencing.

Single-cell sequencing has revolutionized our ability to dissect the heterogeneity within tumor populations. In this study, we present LoRA-TV (Low Rank Approximation with Total Variation), a novel method for clustering tumor cells based on the read depth profiles derived from single-cell sequencing data. Traditional analysis pipelines process read depth profiles of each cell individually. By aggregating shared genomic signatures distributed among individual cells using low-rank optimization and robust smoothing, the proposed method enhances clustering performance. Results from analyses of both simulated and real data demonstrate its effectiveness compared with state-of-the-art alternatives, as supported by improvements in the adjusted Rand index and computational efficiency.

irGSEA: the integration of single-cell rank-based gene set enrichment analysis.

irGSEA is an R package designed to assess the outcomes of various gene set scoring methods when applied to single-cell RNA sequencing data. This package incorporates six distinct scoring methods that rely on the expression ranks of genes, emphasizing relative expression levels over absolute values. The implemented methods include AUCell, UCell, singscore, ssGSEA, JASMINE and Viper. Previous studies have demonstrated the robustness of these methods to variations in dataset size and composition, generating enrichment scores based solely on the relative gene expression of individual cells. By employing the robust rank aggregation algorithm, irGSEA amalgamates results from all six methods to ascertain the statistical significance of target gene sets across diverse scoring methods. The package prioritizes user-friendliness, allowing direct input of expression matrices or seamless interaction with Seurat objects. Furthermore, it facilitates a comprehensive visualization of results. The irGSEA package and its accompanying documentation are accessible on GitHub (https://github.com/chuiqin/irGSEA).

Machine learning of spectra-property relationship for imperfect and small chemistry data.

Machine learning (ML) is causing profound changes to chemical research through its powerful statistical and mathematical methodological capabilities. However, the nature of chemistry experiments often sets very high hurdles to collect high-quality data that are deficiency free, contradicting the need of ML to learn from big data. Even worse, the black-box nature of most ML methods requires more abundant data to ensure good transferability. Herein, we combine physics-based spectral descriptors with a symbolic regression method to establish interpretable spectra-property relationship. Using the machine-learned mathematical formulas, we have predicted the adsorption energy and charge transfer of the CO-adsorbed Cu-based MOF systems from their infrared and Raman spectra. The explicit prediction models are robust, allowing them to be transferrable to small and low-quality dataset containing partial errors. Surprisingly, they can be used to identify and clean error data, which are common data scenarios in real experiments. Such robust learning protocol will significantly enhance the applicability of machine-learned spectroscopy for chemical science.

Fluctuotaxis: Nanoscale directional motion away from regions of fluctuation.

Regulating the motion of nanoscale objects on a solid surface is vital for a broad range of technologies such as nanotechnology, biotechnology, and mechanotechnology. In spite of impressive advances achieved in the field, there is still a lack of a robust mechanism which can operate under a wide range of situations and in a controllable manner. Here, we report a mechanism capable of controllably driving directed motion of any nanoobjects (e.g., nanoparticles, biomolecules, etc.) in both solid and liquid forms. We show via molecular dynamics simulations that a nanoobject would move preferentially away from the fluctuating region of an underlying substrate, a phenomenon termed fluctuotaxis-for which the driving force originates from the difference in atomic fluctuations of the substrate behind and ahead of the object. In particular, we find that the driving force can depend quadratically on both the amplitude and frequency of the substrate and can thus be tuned flexibly. The proposed driving mechanism provides a robust and controllable way for nanoscale mass delivery and has potential in various applications including nanomotors, molecular machines, etc.

Robust ring-opening reaction via asymmetrically coordinated Fe single atoms scaffolded by spoke-like mesoporous carbon nanospheres.

The ability to construct metal single-atom catalysts (SACs) asymmetrically coordinated with organic heteroatoms represents an important endeavor toward developing high-performance catalysts over symmetrically coordinated counterparts. Moreover, it is of key importance in creating supporting matrix with porous architecture for situating SACs as it greatly impacts the mass diffusion and transport of electrolyte. Herein, we report the crafting of Fe single atoms with asymmetrically coordinated nitrogen (N) and phosphorus (P) atoms scaffolded by rationally designed mesoporous carbon nanospheres (MCNs) with spoke-like nanochannels for boosting ring-opening reaction of epoxide to produce an array of pharmacologically important ß-amino alcohols. Notably, interfacial defects in MCN derived from the use of sacrificial template create abundant unpaired electrons, thereby stably anchoring N and P atoms and in turn Fe atoms on MCN. Importantly, the introduction of P atom promotes the symmetry-breaking of common four N-coordinated Fe sites, resulting in the Fe-N3P sites on MCN (denoted Fe-N3P-MCN) with an asymmetric electronic configuration and thus superior catalytic capability. As such, the Fe-N3P-MCN catalysts manifest a high catalytic activity for ring-opening reaction of epoxide (97% yield) over the Fe-N3P docked on nonporous carbon surface (91%) as well as the sole Fe-N4 SACs grounded on the same MCN support (89%). Density functional theory calculations reveal that Fe-N3P SAC lowers the activation barrier for the C-O bond cleavage and the C-N bond formation, thus accelerating the ring-opening of epoxide. Our study provides fundamental and practical insights into developing advanced catalysts in a simple and controllable manner for multistep organic reactions.

Using clustering of genetic variants in Mendelian randomization to interrogate the causal pathways underlying multimorbidity from a common risk factor.

Mendelian randomization (MR) is an epidemiological approach that utilizes genetic variants as instrumental variables to estimate the causal effect of an exposure on a health outcome. This paper investigates an MR scenario in which genetic variants aggregate into clusters that identify heterogeneous causal effects. Such variant clusters are likely to emerge if they affect the exposure and outcome via distinct biological pathways. In the multi-outcome MR framework, where a shared exposure causally impacts several disease outcomes simultaneously, these variant clusters can provide insights into the common disease-causing mechanisms underpinning the co-occurrence of multiple long-term conditions, a phenomenon known as multimorbidity. To identify such variant clusters, we adapt the general method of agglomerative hierarchical clustering to multi-sample summary-data MR setup, enabling cluster detection based on variant-specific ratio estimates. Particularly, we tailor the method for multi-outcome MR to aid in elucidating the causal pathways through which a common risk factor contributes to multiple morbidities. We show in simulations that our "MR-AHC" method detects clusters with high accuracy, outperforming the existing methods. We apply the method to investigate the causal effects of high body fat percentage on type 2 diabetes and osteoarthritis, uncovering interconnected cellular processes underlying this multimorbid disease pair.

From means to meaning in the study of sex/gender differences and similarities.

The incorporation of sex and gender (S/G) related factors is commonly acknowledged as a necessary step to advance towards more personalized diagnoses and treatments for somatic, psychiatric, and neurological diseases. Until now, most attempts to integrate S/G-related factors have been reduced to identifying average differences between females and males in behavioral/ biological variables. The present commentary questions this traditional approach by highlighting three main sets of limitations: 1) Issues stemming from the use of classic parametric methods to compare means; 2) challenges related to the ability of means to accurately represent the data within groups and differences between groups; 3) mean comparisons impose a results' binarization and a binary theoretical framework that precludes advancing towards precision medicine. Alternative methods free of these limitations are also discussed. We hope these arguments will contribute to reflecting on how research on S/G factors is conducted and could be improved.

HBV-infected hepatocellular carcinoma can be robustly classified into three clinically relevant subgroups by a novel analytical protocol.

Liver cancer is the third leading cause of cancer-related death worldwide, and hepatocellular carcinoma (HCC) accounts for a relatively large proportion of all primary liver malignancies. Among the several known risk factors, hepatitis B virus (HBV) infection is one of the important causes of HCC. In this study, we demonstrated that the HBV-infected HCC patients could be robustly classified into three clinically relevant subgroups, i.e. Cluster1, Cluster2 and Cluster3, based on consistent differentially expressed mRNAs and proteins, which showed better generalization. The proposed three subgroups showed different molecular characteristics, immune microenvironment and prognostic survival characteristics. The Cluster1 subgroup had near-normal levels of metabolism-related proteins, low proliferation activity and good immune infiltration, which were associated with its good liver function, smaller tumor size, good prognosis, low alpha-fetoprotein (AFP) levels and lower clinical stage. In contrast, the Cluster3 subgroup had the lowest levels of metabolism-related proteins, which corresponded with its severe liver dysfunction. Also, high proliferation activity and poor immune microenvironment in Cluster3 subgroup were associated with its poor prognosis, larger tumor size, high AFP levels, high incidence of tumor thrombus and higher clinical stage. The characteristics of the Cluster2 subgroup were between the Cluster1 and Cluster3 groups. In addition, MCM2-7, RFC2-5, MSH2, MSH6, SMC2, SMC4, NCPAG and TOP2A proteins were significantly upregulated in the Cluster3 subgroup. Meanwhile, abnormally high phosphorylation levels of these proteins were associated with high levels of DNA repair, telomere maintenance and proliferative features. Therefore, these proteins could be identified as potential diagnostic and prognostic markers. In general, our research provided a novel analytical protocol and insights for the robust classification, treatment and prevention of HBV-infected HCC.

Identifying potential small molecule-miRNA associations via Robust PCA based on γ-norm regularization.

Dysregulation of microRNAs (miRNAs) is closely associated with refractory human diseases, and the identification of potential associations between small molecule (SM) drugs and miRNAs can provide valuable insights for clinical treatment. Existing computational techniques for inferring potential associations suffer from limitations in terms of accuracy and efficiency. To address these challenges, we devise a novel predictive model called RPCA$\Gamma $NR, in which we propose a new Robust principal component analysis (PCA) framework based on $\gamma $-norm and $l_{2,1}$-norm regularization and design an Augmented Lagrange Multiplier method to optimize it, thereby deriving the association scores. The Gaussian Interaction Profile Kernel Similarity is calculated to capture the similarity information of SMs and miRNAs in known associations. Through extensive evaluation, including Cross Validation Experiments, Independent Validation Experiment, Efficiency Analysis, Ablation Experiment, Matrix Sparsity Analysis, and Case Studies, RPCA$\Gamma $NR outperforms state-of-the-art models concerning accuracy, efficiency and robustness. In conclusion, RPCA$\Gamma $NR can significantly streamline the process of determining SM-miRNA associations, thus contributing to advancements in drug development and disease treatment.

Universal structural requirements for maximal robust perfect adaptation in biomolecular networks.

Adaptation is a running theme in biology. It allows a living system to survive and thrive in the face of unpredictable environments by maintaining key physiological variables at their desired levels through tight regulation. When one such variable is maintained at a certain value at the steady state despite perturbations to a single input, this property is called robust perfect adaptation (RPA). Here we address and solve the fundamental problem of maximal RPA (maxRPA), whereby, for a designated output variable, RPA is achieved with respect to perturbations in virtually all network parameters. In particular, we show that the maxRPA property imposes certain structural constraints on the network. We then prove that these constraints are fully characterized by simple linear algebraic stoichiometric conditions which differ between deterministic and stochastic descriptions of the dynamics. We use our results to derive a new internal model principle (IMP) for biomolecular maxRPA networks, akin to the celebrated IMP in control theory. We exemplify our results through several known biological examples of robustly adapting networks and construct examples of such networks with the aid of our linear algebraic characterization. Our results reveal the universal requirements for maxRPA in all biological systems, and establish a foundation for studying adaptation in general biomolecular networks, with important implications for both systems and synthetic biology.

A genetic mammalian proportional-integral feedback control circuit for robust and precise gene regulation.

The processes that keep a cell alive are constantly challenged by unpredictable changes in its environment. Cells manage to counteract these changes by employing sophisticated regulatory strategies that maintain a steady internal milieu. Recently, the antithetic integral feedback motif has been demonstrated to be a minimal and universal biological regulatory strategy that can guarantee robust perfect adaptation for noisy gene regulatory networks in Escherichia coli. Here, we present a realization of the antithetic integral feedback motif in a synthetic gene circuit in mammalian cells. We show that the motif robustly maintains the expression of a synthetic transcription factor at tunable levels even when it is perturbed by increased degradation or its interaction network structure is perturbed by a negative feedback loop with an RNA-binding protein. We further demonstrate an improved regulatory strategy by augmenting the antithetic integral motif with additional negative feedback to realize antithetic proportional-integral control. We show that this motif produces robust perfect adaptation while also reducing the variance of the regulated synthetic transcription factor. We demonstrate that the integral and proportional-integral feedback motifs can mitigate the impact of gene expression burden, and we computationally explore their use in cell therapy. We believe that the engineering of precise and robust perfect adaptation will enable substantial advances in industrial biotechnology and cell-based therapeutics.

Boosting edgeR (Robust) by dealing with missing observations and gene-specific outliers in RNA-Seq profiles and its application to explore biomarker genes for diagnosis and therapies of ovarian cancer.

The edgeR (Robust) is a popular approach for identifying differentially expressed genes (DEGs) from RNA-Seq profiles. However, it shows weak performance against gene-specific outliers and is unable to handle missing observations. To address these issues, we proposed a pre-processing approach of RNA-Seq count data by combining the iLOO-based outlier detection and random forest-based missing imputation approach for boosting the performance of edgeR (Robust). Both simulation and real RNA-Seq count data analysis results showed that the proposed edgeR (Robust) outperformed than the conventional edgeR (Robust). To investigate the effectiveness of identified DEGs for diagnosis, and therapies of ovarian cancer (OC), we selected top-ranked 12 DEGs (IL6, XCL1, CXCL8, C1QC, C1QB, SNAI2, TYROBP, COL1A2, SNAP25, NTS, CXCL2, and AGT) and suggested hub-DEGs guided top-ranked 10 candidate drug-molecules for the treatment against OC. Hence, our proposed procedure might be an effective computational tool for exploring potential DEGs from RNA-Seq profiles for diagnosis and therapies of any disease.

Highly Elastic, Robust, and Efficient Hydrogel Solar Absorber against Harsh Environmental Impacts.

Solar distillation is a promising approach for addressing water scarcity, but relentless stress/strain perturbations induced by wind and waves would inevitably cause structural damage to solar absorbers. Despite notable advances in efficient solar absorbers, there have been no reports of compliant and robust solar absorbers withstanding practical mechanical impacts. Herein, an elastic and robust hydrogel absorber that exhibited a high level of evaporation performance was fabricated by introducing ion-coordinated MXene nanosheets as photothermal conversion units and mechanically enhanced fillers. The ion-coordinated MXene nanosheets acting as strong cross-linking points provided excellent elasticity and robustness to the hydrogel absorber. As a result, the evaporation rate of hydrogel absorber, with a high initial value of 2.61 kg m-2 h-1 under one sun irradiation, remained at 2.15 kg m-2 h-1 under a 100% tensile strain state and 2.40 kg m-2 h-1 after 10 000 stretching-releasing cycles. This continuous and stable water desalination approach provides a promising device for actual seawater distillation.

Ultrafast Ion Transfer of Metal-Organic Framework Interface for Highly Efficient Energy Storage.

Flexible supercapacitors are favorable for wearable electronics. However, their high-rate capability and mechanical properties are limited because of unsatisfactory ion transfer kinetics and interfacial modulus mismatch inside devices. Here, we develop a metal-organic framework interface with superior electrical and mechanical properties for supercapacitors. The interfacial mechanism facilitates ultrafast ion transfer with an energy barrier reduction of 43% compared with that of conventional transmembrane transport. It delivers high specific capacity at a wide rate range and exhibits ultrastability beyond 30000 charge-discharge cycles. Furthermore, meliorative modulus mismatch benefited from ultrathin interface design that improves mechanical properties of flexible supercapacitors. It delivers a stable energy supply under various mechanical conditions like bending and twisting status and displays ultrastable mechanical properties with performance retention of 95.5% after 10 000 bending cycles. The research paves the way for interfacial engineering for ultrastable electrochemical devices.

Robust, Precise, and Deep Proteome Profiling Using a Small Mass Range and Narrow Window Data-Independent-Acquisition Scheme.

In recent years, a plethora of different data-independent acquisition methods have been developed for proteomics to cover a wide range of requirements. Current deep proteome profiling methods rely on fractionations, elaborate chromatography, and mass spectrometry setups or display suboptimal quantitative precision. We set out to develop an easy-to-use one shot DIA method that achieves high quantitative precision and high proteome coverage. We achieve this by focusing on a small mass range of 430-670 m/z using small isolation windows without overlap. With this new method, we were able to quantify >9200 protein groups in HEK lysates with an average coefficient of variance of 3.2%. To demonstrate the power of our newly developed narrow mass range method, we applied it to investigate the effect of PGC-1α knockout on the skeletal muscle proteome in mice. Compared to a standard data-dependent acquisition method, we could double proteome coverage and, most importantly, achieve a significantly higher quantitative precision, as compared to a previously proposed DIA method. We believe that our method will be especially helpful in quantifying low abundant proteins in samples with a high dynamic range. All raw and result files are available at massive.ucsd.edu (MSV000092186).

RoPE: A robust profile likelihood method for differential gene expression analysis.

Variation in RNA-Seq data creates modeling challenges for differential gene expression (DE) analysis. Statistical approaches address conventional small sample sizes and implement empirical Bayes or non-parametric tests, but frequently produce different conclusions. Increasing sample sizes enable proposal of alternative DE paradigms. Here we develop RoPE, which uses a data-driven adjustment for variation and a robust profile likelihood ratio DE test. Simulation studies show RoPE can have improved performance over existing tools as sample size increases and has the most reliable control of error rates. Application of RoPE demonstrates that an active Pseudomonas aeruginosa infection downregulates the SLC9A3 Cystic Fibrosis modifier gene.

Genome-wide association study of multiyear dynamic growth traits in hybrid Liriodendron identifies robust genetic loci associated with growth trajectories.

The genetic factors underlying growth traits differ over time points or stages. However, most current studies of phenotypes at single time points do not capture all loci or explain the genetic differences underlying growth trajectories. Hybrid Liriodendron exhibits obvious heterosis and is widely cultivated, although its complex genetic mechanism underlying growth traits remains unknown. A genome-wide association study (GWAS) is an effective method for elucidating the genetic architecture by identifying genetic loci underlying complex quantitative traits. In the present study, using a GWAS, we identified robust loci associated with growth trajectories in hybrid Liriodendron populations. We selected 233 hybrid progenies derived from 25 crosses for resequencing, and measured their tree height (H) and diameter at breast height (DBH) for 11 consecutive years; 192 972 high-quality single nucleotide polymorphisms (SNPs) were obtained. The dynamics of the multiyear single-trait GWAS showed that year-specific SNPs predominated, and only five robust SNPs for DBH were identified in at least three different years. Multitrait GWAS analysis with model parameters as latent variables also revealed 62 SNPs for H and 52 for DBH associated with the growth trajectory, displaying different biomass accumulation patterns, among which four SNPs exerted pleiotropic effects. All identified SNPs also exhibited temporal variations in effect sizes and inheritance patterns potentially related to different growth and developmental stages. The haplotypes resulting from these significant SNPs might pyramid favorable loci, benefitting the selection of superior genotypes. The present study provides insights into the genetic architecture of dynamic growth traits and lays a basis for future molecular-assisted breeding.

Learning Optimal Dynamic Treatment Regimes from Longitudinal Data.

Studies often report estimates of the average treatment effect (ATE). While the ATE summarizes the effect of a treatment on average, it does not provide any information about the effect of treatment within any individual. A treatment strategy that uses an individual's information to tailor treatment to maximize benefit is known as an optimal dynamic treatment rule (ODTR). Treatment, however, is typically not limited to a single point in time; consequently, learning an optimal rule for a time-varying treatment may involve not just learning the extent to which the comparative treatments' benefits vary across the characteristics of individuals, but also learning the extent to which the comparative treatments' benefits vary as relevant circumstances evolve within an individual. The goal of this paper is to provide a tutorial for estimating ODTR from longitudinal observational and clinical trial data for applied researchers. We describe an approach that uses a doubly-robust unbiased transformation of the conditional average treatment effect. We then learn a time-varying ODTR for when to increase buprenorphine-naloxone (BUP-NX) dose to minimize return-to-regular-opioid-use among patients with opioid use disorder. Our analysis highlights the utility of ODTRs in the context of sequential decision making: the learned ODTR outperforms a clinically defined strategy.

Extending prediction models for use in a new target population with failure time outcomes.

Prediction models are often built and evaluated using data from a population that differs from the target population where model-derived predictions are intended to be used in. In this article, we present methods for evaluating model performance in the target population when some observations are right censored. The methods assume that outcome and covariate data are available from a source population used for model development and covariates, but no outcome data, are available from the target population. We evaluate the finite sample performance of the proposed estimators using simulations and apply the methods to transport a prediction model built using data from a lung cancer screening trial to a nationally representative population of participants eligible for lung cancer screening.