Online Data Repositories Research Articles

Environmental footprints of the data center service sector in Sweden

The global data center (DC) sector has expanded rapidly during the last decades, due to the rising demand for digital services. In the Nordic region, Sweden has emerged as a global hub, attracting leading technology companies like Amazon, Facebook, Microsoft, and Google. Server halls of DCs are energy intensive buildings, which puts pressure on local water resources and contributes to global greenhouse gas emissions. This study aims to, firstly, quantify the environmental impact of DCs, based on energy usage, water consumption, and greenhouse gas (GHG) emissions. Secondly, it develops a planning tool by employing a multi-criteria approach to optimally locate new DCs and to assess the site suitability of existing ones in Sweden. Data of various performance indicators (geographical data on renewable energy accessibility, free cooling conditions, excess heat receivers, and resilience to water shortages) of DCs was collected through different means, e.g., questionnaire surveys, permit applications, company websites, and other open online data repositories. ArcGIS Pro was employed for spatial analysis, and 68 DCs with a site suitability index (SSI) ≤ 45 % were identified as less ideally located. The principal findings are centered on Sweden, and thereby primarily benefit stakeholders engaged in decision-making for evaluating existing or strategic planning of new DCs by incorporating a comprehensive environmental perspective. Given the rapidly changing climate, strategically siting DCs will become crucial for minimizing the sector's environmental impact.

Eliminating invisible deaths: the woeful state of global rabies data and its impact on progress towards 2030 sustainable development goals for neglected tropical diseases

Like other neglected diseases, surveillance data for rabies is insufficient and incompatible with the need to accurately describe the burden of disease. Multiple modeling studies central to estimating global human rabies deaths have been conducted in the last two decades, with results ranging from 14,000 to 74,000 deaths annually. Yet, uncertainty in model parameters, inconsistency in modeling approaches, and discrepancies in data quality per country included in global burden studies have led to recent skepticism about the magnitude of rabies mortality. Lack of data not only limits the efficiency and monitoring of rabies elimination strategies but also severely diminishes abilities to advocate for support from international funding agencies. Meanwhile, the most vulnerable communities continue to suffer from deaths that could have been prevented through more robust reporting. The Zero by 30 global strategy to eliminate dog-mediated human rabies by 2030 recommends endemic countries adopt the intersectoral approach, Integrated Bite Case Management (IBCM), as a cost-effective method to enhance surveillance. However, effective implementation of IBCM is impeded by challenges such as limited capacity, resources, knowledge, skills, and attitudes toward compliance. To address this, the World Health Organization and United Against Rabies Forum have developed several open-access tools to guide national control programs in strong data collection practices, and online data repositories to pragmatically streamline reporting and encourage data sharing. Here, we discuss how current and future initiatives can be best employed to improve the implementation of existing surveillance tools and prioritization of effective data reporting/sharing to optimize progress toward 2030 elimination.

Teachers' judgment accuracy: A replication check by psychometric meta-analysis.

Teachers' judgment accuracy is a core competency in their daily business. Due to its importance, several meta-analyses have estimated how accurately teachers judge students' academic achievements by measuring teachers' judgment accuracy (i.e., the correlation between teachers' judgments of students' academic abilities and students' scores on achievement tests). In our study, we considered previous meta-analyses and updated these databases and the analytic combination of data using a psychometric meta-analysis to explain variations in results across studies. Our results demonstrate the importance of considering aggregation and publication bias as well as correcting for the most important artifacts (e.g., sampling and measurement error), but also that most studies fail to report the data needed for conducting a meta-analysis according to current best practices. We find that previous reviews have underestimated teachers' judgment accuracy and overestimated the variance in estimates of teachers' judgment accuracy across studies because at least 10% of this variance may be associated with common artifacts. We conclude that ignoring artifacts, as in classical meta-analysis, may lead one to erroneously conclude that moderator variables, instead of artifacts, explain any variation. We describe how online data repositories could improve the scientific process and the potential for using psychometric meta-analysis to synthesize results and assess replicability.

Alismataceae in humid subtropical areas in southern Brazil

Abstract Alismataceae is a family of aquatic or semi-aquatic plants with a wide distribution in humid areas, occurring in tropical, subtropical and temperate regions. Although Brazil accounts for 50% of the entire species diversity of this family, there are many gaps in knowledge of the species and their distribution, especially in the southern region of the country. This paper presents the study of Alismataceae species in the Biomes Pampa and Atlantic Forest of Rio Grande do Sul, highlighting taxonomic, ecological and distribution aspects. This study was based on the morphological analysis of specimens deposited in online data repositories and herbaria, as well as specimens collected in wetlands in the extreme south of Brazil. Ten taxa were registered for Rio Grande do Sul state: Echinodorus grandiflorus, E. longiscapus, E. macrophyllus, E. reptilis, E. uruguayensis, Helanthium tenellum, Hydrocleys nymphoides, Sagittaria lancifolia, S. rhombifolia and S. montevidensis. As a result of this study, an identification key for the taxa, morphological descriptions, images of the species and distribution maps are presented. Ecological and phenological data are also provided.

Identification and Characterization of ten Escherichia coli Strains Encoding Novel Shiga Toxin 2 Subtypes, Stx2n as Well as Stx2j, Stx2m, and Stx2o, in the United States.

The sharing of genome sequences in online data repositories allows for large scale analyses of specific genes or gene families. This can result in the detection of novel gene subtypes as well as the development of improved detection methods. Here, we used publicly available WGS data to detect a novel Stx subtype, Stx2n in two clinical E. coli strains isolated in the USA. During this process, additional Stx2 subtypes were detected; six Stx2j, one Stx2m strain, and one Stx2o, were all analyzed for variability from the originally described subtypes. Complete genome sequences were assembled from short- or long-read sequencing and analyzed for serotype, and ST types. The WGS data from Stx2n- and Stx2o-producing STEC strains were further analyzed for virulence genes pro-phage analysis and phage insertion sites. Nucleotide and amino acid maximum parsimony trees showed expected clustering of the previously described subtypes and a clear separation of the novel Stx2n subtype. WGS data were used to design OMNI PCR primers for the detection of all known stx1 (283 bp amplicon), stx2 (400 bp amplicon), intimin encoded by eae (221 bp amplicon), and stx2f (438 bp amplicon) subtypes. These primers were tested in three different laboratories, using standard reference strains. An analysis of the complete genome sequence showed variability in serogroup, virulence genes, and ST type, and Stx2 pro-phages showed variability in size, gene composition, and phage insertion sites. The strains with Stx2j, Stx2m, Stx2n, and Stx2o showed toxicity to Vero cells. Stx2j carrying strain, 2012C-4221, was induced when grown with sub-inhibitory concentrations of ciprofloxacin, and toxicity was detected. Taken together, these data highlight the need to reinforce genomic surveillance to identify the emergence of potential new Stx2 or Stx1 variants. The importance of this surveillance has a paramount impact on public health. Per our description in this study, we suggest that 2017C-4317 be designated as the Stx2n type-strain.

Comparing containerization-based approaches for reproducible computational modeling of environmental systems

Creating online data repositories that follow findable, accessible, interoperable, and reusable (FAIR) principles has been a significant focus in the research community to address the reproducibility crisis facing many computational fields, including environmental modeling. However, less work has focused on another reproducibility challenge: capturing modeling software and computational environments needed to reproduce complex modeling workflows. Containerization technology offers an opportunity to address this need, and there are a growing number of strategies being put forth that leverage containerization to improve the reproducibility of environmental modeling. This research compares ten such approaches using a hydrologic model application as a case study. For each approach, we use both quantitative and qualitative metrics for comparing the different strategies. Based on the results, we discuss challenges and opportunities for containerization in environmental modelling and recommend best practices across both research and educational use cases for when and how to apply the different containerization-based strategies.

MESSES: Software for Transforming Messy Research Datasets into Clean Submissions to Metabolomics Workbench for Public Sharing.

In recent years, the FAIR guiding principles and the broader concept of open science has grown in importance in academic research, especially as funding entities have aggressively promoted public sharing of research products. Key to public research sharing is deposition of datasets into online data repositories, but it can be a chore to transform messy unstructured data into the forms required by these repositories. To help generate Metabolomics Workbench depositions, we have developed the MESSES (Metadata from Experimental SpreadSheets Extraction System) software package, implemented in the Python 3 programming language and supported on Linux, Windows, and Mac operating systems. MESSES helps transform tabular data from multiple sources into a Metabolomics Workbench specific deposition format. The package provides three commands, extract, validate, and convert, that implement a natural data transformation workflow. Moreover, MESSES facilitates richer metadata capture than is typically attempted by manual efforts. The source code and extensive documentation is hosted on GitHub and is also available on the Python Package Index for easy installation.

Surprising implications of differences in locations versus differences in means

Social science researchers depend on differences in means between experimental and control conditions to draw substantive conclusions. However, an alternative is to use differences in locations. For normal distributions, means and locations are the same, but for skew normal distributions, means and locations are different. If a difference in means and locations are similar, and in the same direction, the resulting substantive story may be similar. However, if a difference in means and locations are dissimilar, especially if they oppose directionally, the resulting substantive story may differ dramatically. We collected 51 data sets from online data repositories to check how often the differences in means versus locations are substantially different or are in different directions. Although the values depend on what one counts, the overall conclusion is that the two types of differences have a larger than trivial chance of disagreeing substantially. We suggest that when researchers report normal statistics (mean and standard deviation), they should report skew normal statistics (location, scale, and shape) too, against the nontrivial chance that the skew normal statistics imply a substantive story in opposition to that implied by the normal statistics.

Distance and Fieldwork in a Pandemic: How Not ‘Being there’ is Impacting Research on Japan

Fieldwork is about ‘being there’ in the field to gather data. But what happens when researchers cannot visit the field? This article explores how Japan scholars have been dealing with the impact of COVID-19 on their research. It examines how restrictions on travel affect access to materials and engagement with fieldwork subjects, highlighting how the pandemic has created both obstructions and opportunities. The term ‘fieldwork’ usually involves ethnographic methods of data collection such as participant observation and interviews, but a wider interpretation encompasses visiting archives, libraries and museums. This article takes an inclusive definition of fieldwork to discover the impact of not ‘being there’ for scholars of Japan. Findings show that the inability to ‘be there’ has led more scholars to seek out material from online data repositories, archives and library collections. However, while the demand for online resources is increasing, materials are not always easily accessible to Japan scholars. The impact of travel restrictions on librarians has in turn affected the aggregation of materials, which has occasioned scholars to seek alternative methods of sourcing materials. Finally, through a case study, this article examines the methods scholars are adopting in digital ethnographic data collection to adjust to not ‘being there’.

Science paper or big data? Assessing invasion dynamics using observational data

Non-native species are spreading at an unprecedented rate over large spatial scales, with global environmental change and growth in commerce providing novel opportunities for range expansion. Assessing the pattern and rate of spread is key to the development of strategies for safeguarding against future invasions and efficiently managing existing ones. Such assessments often depend on spatial distribution data from online repositories, which can be spatially biased, imprecise, and lacking in quantity. Here, the influence of disparities between occurrence records from online data repositories and what is known of the invasion history from peer-reviewed published literature on non-native species range expansion was evaluated using 6693 records of the Pacific oyster, Magallana gigas (Thunberg, 1793), spanning 56 years of its invasion in Europe. Two measures of spread were calculated: maximum rate of spread (distance from introduction site over time) and accumulated area (spatial expansion). Results suggest that despite discrepancies between online and peer-reviewed data sources, including a paucity of records from the early invasion history in online repositories, the use of either source does not result in significantly different estimates of spread. Our study significantly improves our understanding of the European distribution of M. gigas and suggests that a combination of short- and long-range dispersal drives range expansions. More widely, our approach provides a framework for comparison of online occurrence records and invasion histories as documented in the peer-reviewed literature, allowing critical evaluation of both data sources and improving our understanding of invasion dynamics significantly.

The diversity of cultivable endophytic fungi of the sand coast plant Ipomoeapes-caprae in Taiwan.

Ipomoeapes-caprae is a plant of sand coasts and it can tolerate stresses, such as high salinity, strong wind and sand movements and lack of nutrients. It plays an important role in coast protection and preventing erosion. Fungal endophytes show high biodiversity and have a strong influence on the survival of plants under different stress factors. Although this plant is important for sand coast ecosystems, little is known about the associated fungi. In this study, we isolated and identified endophytic fungi of Ipomoeapes-caprae, a dominant plant along the shore of Taiwan. The dataset contains 896 records, which correspond to 177 species. The geographical scope of the dataset covers the northern subtropical area of the main island of Taiwan, with its sand coasts in New Taipei, Taoyuan, Hsinchu and Taichung and two botanical gardens in Taipei and Taichung. The detailed original data of fungal diversity are rarely publicly shared under strictly formalised and, thus, reusable standards. As an example for such an approach, the complete occurrence dataset was made available in the Darwin Core Archive format via the Global Biodiversity Information Facility (GBIF) under Version 1.13, Taiwan Biodiversity Information Facility (TaiBIF) https://doi.org/10.15468/9h9rcg. In this first data paper on endophytic fungi, the scientific name and associated DNA sequence in the dataset were directly linked to other free online resource (Index Fungorum, GenBank), which shows the potential of GBIF for linking together different online data repositories. We describe a dataset, in which the diversity of endophytic fungi of the sand coast plant Ipomoeapes-caprae in Taiwan was investigated.

Sampling from commercial vessel routes can capture marine biodiversity distributions effectively.

Collecting fine-scale occurrence data for marine species across large spatial scales is logistically challenging but is important to determine species distributions and for conservation planning. Inaccurate descriptions of species ranges could result in designating protected areas with inappropriate locations or boundaries. Optimizing sampling strategies therefore is a priority for scaling up survey approaches using tools such as environmental DNA (eDNA) to capture species distributions. In a marine context, commercial vessels, such as ferries, could provide sampling platforms allowing access to undersampled areas and repeatable sampling over time to track community changes. However, sample collection from commercial vessels could be biased and may not represent biological and environmental variability. Here, we evaluate whether sampling along Mediterranean ferry routes can yield unbiased biodiversity survey outcomes, based on perfect knowledge from a stacked species distribution model (SSDM) of marine megafauna derived from online data repositories. Simulations to allocate sampling point locations were carried out representing different sampling strategies (random vs regular), frames (ferry routes vs unconstrained), and number of sampling points. SSDMs were remade from different sampling simulations and compared with the "perfect knowledge" SSDM to quantify the bias associated with different sampling strategies. Ferry routes detected more species and were able to recover known patterns in species richness at smaller sample sizes better than unconstrained sampling points. However, to minimize potential bias, ferry routes should be chosen to cover the variability in species composition and its environmental predictors in the SSDMs. The workflow presented here can be used to design effective sampling strategies using commercial vessel routes globally for eDNA and other biodiversity survey techniques. This approach has potential to provide a cost-effective method to access remote oceanic areas on a regular basis and can recover meaningful data on spatiotemporal biodiversity patterns.

Unified and Standardized Mass Spectrometry Data Processing in Python Using spectrum_utils.

spectrum_utils is a Python package for mass spectrometry data processing and visualization. Since its introduction, spectrum_utils has grown into a fundamental software solution that powers various applications in proteomics and metabolomics, ranging from spectrum preprocessing prior to spectrum identification and machine learning applications to spectrum plotting from online data repositories and assisting data analysis tasks for dozens of other projects. Here, we present updates to spectrum_utils, which include new functionality to integrate mass spectrometry community data standards, enhanced mass spectral data processing, and unified mass spectral data visualization in Python. spectrum_utils is freely available as open source at https://github.com/bittremieux/spectrum_utils.

Current State of Microplastic Pollution Research Data: Trends in Availability and Sources of Open Data

The rapid growth in microplastic pollution research is influencing funding priorities, environmental policy, and public perceptions of risks to water quality and environmental and human health. Ensuring that environmental microplastics research data are findable, accessible, interoperable, and reusable (FAIR) is essential to inform policy and mitigation strategies. We present a bibliographic analysis of data sharing practices in the environmental microplastics research community, highlighting the state of openness of microplastics data. A stratified (by year) random subset of 785 of 6,608 microplastics articles indexed in Web of Science indicates that, since 2006, less than a third (28.5%) contained a data sharing statement. These statements further show that most often, the data were provided in the articles’ supplementary material (38.8%) and only 13.8% via a data repository. Of the 279 microplastics datasets found in online data repositories, 20.4% presented only metadata with access to the data requiring additional approval. Although increasing, the rate of microplastic data sharing still lags behind that of publication of peer-reviewed articles on environmental microplastics. About a quarter of the repository data originated from North America (12.8%) and Europe (13.4%). Marine and estuarine environments are the most frequently sampled systems (26.2%); sediments (18.8%) and water (15.3%) are the predominant media. Of the available datasets accessible, 15.4% and 18.2% do not have adequate metadata to determine the sampling location and media type, respectively. We discuss five recommendations to strengthen data sharing practices in the environmental microplastic research community.

AN EXPLORATION OF BARRIERS TO OFFENDER REINTEGRATION: PROBATION AND PRISON OFFICER OPINIONS VS PUBLIC OPINION

Stigmatization has been a major hindrance to successful reintegration of released offenders into the community primarily due to increased ostracism and anxiety of the wider community relative to recidivism and public security. In addition, the fear of reoffending has been extended to offender employability, housing, and resumption to learning institutions, as well as reconnecting with the community for personal and social growth. Consequently, probation and prison officers initiate different programs geared towards social reintegration to facilitate apposite transitioning of released offenders back to the community. This study, therefore, evaluated the position of the general public and probation and prison officers in driving successful reintegration of offenders by assessing their opinions and perceptions towards social reintegration to isolate barriers to successful social reintegration while reducing recidivism. A mixed-methods research guided by a pragmatist epistemological and ontological paradigm was conducted. In the process, a desktop search was carried out on online data repositories and journal databases to gather secondary quantitative and qualitative data, which was subjected to a systematic review. Following descriptive analysis on secondary quantitative data, and thematic analysis on secondary qualitative data, it was established that probation and prison officers opined the need to blend social reintegration programs with community initiatives, especially those that support offender employment and reemployment. Importantly, it was established that stigma is amplified by the potential of the higher number of released offenders to recidivate negatively affecting public safety and security.

Importance of Seasonal Variation in Hawaiian Mushroom (Agaricomycetes) Basidiomata Production for Biodiversity Discovery and Conservation.

The Hawaiian Islands have a relatively well-known funga for a tropical location, yet there are over 400 species of mushrooms (Agaricomycetes) in the archipelago that remain to be documented. Importantly, the International Union for Conservation of Nature (IUCN) recently evaluated six mushrooms endemic to the islands as threatened with extinction. To improve detection of mushrooms for biodiversity discovery and better monitor threatened species in the archipelago-where many localities lack strong annual precipitation patterns associated with an obvious season for increased mushroom basidiomata production-we examined the phenology of Hawaiian mushrooms. Monthly richness was determined from a literature review and abundance from online data repositories. Phenological patterns were separately explored for native species and differing elevation and annual precipitation categories. Despite relatively consistent monthly temperatures and areas with regular monthly rainfall, we found Hawaiian mushrooms generally exhibit uneven temporal patterns in basidiomata production: richness and abundance are generally highest in January and lowest from February to April, then usually increase from May to July and remain at elevated levels through December. This pattern does not occur when considering native species richness only, nor when examining abundance data stratified by elevation and annual rainfall categories. Increased monthly basidiomata abundance in low elevation (<1,000 m), dry (<1,000 mm rainfall/year) locations on O'ahu and low, mesic (1,000-2,500 mm rainfall/year) locations on O'ahu and Kaua'i are positively correlated with increased monthly rainfall. Phenology of macrofungal sporocarp production should potentially be included in species threat assessments by the IUCN to increase detection via traditional surveying methods.

Selective data acquisition in the wild for model charging

The lack of sufficient labeled data is a key bottleneck for practitioners in many real-world supervised machine learning (ML) tasks. In this paper, we study a new problem, namely selective data acquisition in the wild for model charging : given a supervised ML task and data in the wild (e.g., enterprise data warehouses, online data repositories, data markets, and so on), the problem is to select labeled data points from the data in the wild as additional train data that can help the ML task. It consists of two steps (Fig. 1). The first step is to discover relevant datasets ( e.g. , tables with similar relational schema), which will result in a set of candidate datasets. Because these candidate datasets come from different sources and may follow different distributions, not all data points they contain can help. The second step is to select which data points from these candidate datasets should be used. We build an end-to-end solution. For step 1, we piggyback off-the-shelf data discovery tools. Technically, our focus is on step 2, for which we propose a solution framework called AutoData. It first clusters all data points from candidate datasets such that each cluster contains similar data points from different sources. It then iteratively picks which cluster to use, samples data points ( i.e. , a mini-batch) from the picked cluster, evaluates the mini-batch, and then revises the search criteria by learning from the feedback ( i.e. , reward) based on the evaluation. We propose a multi-armed bandit based solution and a Deep Q Networks-based reinforcement learning solution. Experiments using both relational and image datasets show the effectiveness of our solutions.

Effective k-Means Clustering in Greedy Prepruned Tree-based Classification for Obstructive Sleep Apnea

Incorporation of prepruned decision trees to k-means clustering through one to three types of tree-depth controllers and cluster partitioning was done to develop a combined algorithm named as Greedy Pre-pruned Tree-based Clustering (GPrTC) algorithm. Pre-pruned clustered decision trees are applied in a greedy concerted way to five datasets of obstructive sleep apnea and others from online data repositories. The optimal number of k clusters for k-means clustering is determined after trees are greedily prepruned by tree-depth controllers of minimum number of leaf nodes, minimum number of parent nodes and maximum number of tree splitting. After applying the GPrTC algorithm to the assigned datasets, when compared with the conventional k-means clustering, results showed that the former has significantly lower average distortion per point and lower average run-time for 2-D and 3-D data over around 30 thousand points. Classification efficiency and speed of the former algorithm is more than two times better the latter algorithm over a higher range of points being run. GPrTC algorithm showed better classification accuracies than k-means clustering in almost all the assigned datasets. This concludes that the proposed algorithm is significantly much more efficient, less distortion and much faster than k-means clustering with moderately better in terms of classification and/or prediction accuracies.

Systems Biology-Derived Genetic Signatures of Mastitis in Dairy Cattle: A New Avenue for Drug Repurposing.

Simple SummaryTherapeutic success of bovine mastitis depends mainly on accurately diagnosing the type of pathogen involved. Despite the development prospects for bovine mastitis diagnosis, including new biomarker discovery to target specific pathogens with high sensitivity and specificity, treatment studies have shown controversial results, and the most efficient, safe, and economical treatments for mastitis are still topics of scientific debate. The goal of this research is the integration of different levels of systems biology data to predict candidate drugs for the control and management of E. coli mastitis. We propose that the novel drugs could be used by pharmaceutical scientists or veterinarians to find commercially efficacious medicines.Mastitis, a disease with high incidence worldwide, is the most prevalent and costly disease in the dairy industry. Gram-negative bacteria such as Escherichia coli (E. coli) are assumed to be among the leading agents causing acute severe infection with clinical signs. E. Coli, environmental mastitis pathogens, are the primary etiological agents of bovine mastitis in well-managed dairy farms. Response to E. Coli infection has a complex pattern affected by genetic and environmental parameters. On the other hand, the efficacy of antibiotics and/or anti-inflammatory treatment in E. coli mastitis is still a topic of scientific debate, and studies on the treatment of clinical cases show conflicting results. Unraveling the bio-signature of mastitis in dairy cattle can open new avenues for drug repurposing. In the current research, a novel, semi-supervised heterogeneous label propagation algorithm named Heter-LP, which applies both local and global network features for data integration, was used to potentially identify novel therapeutic avenues for the treatment of E. coli mastitis. Online data repositories relevant to known diseases, drugs, and gene targets, along with other specialized biological information for E. coli mastitis, including critical genes with robust bio-signatures, drugs, and related disorders, were used as input data for analysis with the Heter-LP algorithm. Our research identified novel drugs such as Glibenclamide, Ipratropium, Salbutamol, and Carbidopa as possible therapeutics that could be used against E. coli mastitis. Predicted relationships can be used by pharmaceutical scientists or veterinarians to find commercially efficacious medicines or a combination of two or more active compounds to treat this infectious disease.

Interpretable Machine Learning of Chemical Bonding at Solid Surfaces.

Understanding the nature of chemical bonding and its variation in strength across physically tunable factors is important for the development of novel catalytic materials. One way to speed up this process is to employ machine learning (ML) algorithms with online data repositories curated from high-throughput experiments or quantum-chemical simulations. Despite the reasonable predictive performance of ML models for predicting reactivity properties of solid surfaces, the ever-growing complexity of modern algorithms, e.g., deep learning, makes them black boxes with little to no explanation. In this Perspective, we discuss recent advances of interpretable ML for opening up these black boxes from the standpoints of feature engineering, algorithm development, and post hoc analysis. We underline the pivotal role of interpretability as the foundation of next-generation ML algorithms and emerging AI platforms for driving discoveries across scientific disciplines.