A community-powered search of machine learning strategy space to find NMR property prediction models.

The rise of machine learning (ML) has created an explosion in the potential strategies for using data to make scientific predictions. For physical scientists wishing to apply ML strategies to a particular domain, it can be difficult to assess in advance what strategy to adopt within a vast space of possibilities. Here we outline the results of an online community-powered effort to swarm search the space of ML strategies and develop algorithms for predicting atomic-pairwise nuclear magnetic resonance (NMR) properties in molecules. Using an open-source dataset, we worked with Kaggle to design and host a 3-month competition which received 47,800 ML model predictions from 2,700 teams in 84 countries. Within 3 weeks, the Kaggle community produced models with comparable accuracy to our best previously published 'in-house' efforts. A meta-ensemble model constructed as a linear combination of the top predictions has a prediction accuracy which exceeds that of any individual model, 7-19x better than our previous state-of-the-art. The results highlight the potential of transformer architectures for predicting quantum mechanical (QM) molecular properties.

Trends and gaps in the use of citizen science derived data as input for species distribution models: A quantitative review.

Citizen science (CS) currently refers to the participation of non-scientist volunteers in any discipline of conventional scientific research. Over the last two decades, nature-based CS has flourished due to innovative technology, novel devices, and widespread digital platforms used to collect and classify species occurrence data. For scientists, CS offers a low-cost approach of collecting species occurrence information at large spatial scales that otherwise would be prohibitively expensive. We examined the trends and gaps linked to the use of CS as a source of data for species distribution models (SDMs), in order to propose guidelines and highlight solutions. We conducted a quantitative literature review of 207 peer-reviewed articles to measure how the representation of different taxa, regions, and data types have changed in SDM publications since the 2010s. Our review shows that the number of papers using CS for SDMs has increased at approximately double the rate of the overall number of SDM papers. However, disparities in taxonomic and geographic coverage remain in studies using CS. Western Europe and North America were the regions with the most coverage (73%). Papers on birds (49%) and mammals (19.3%) outnumbered other taxa. Among invertebrates, flying insects including Lepidoptera, Odonata and Hymenoptera received the most attention. Discrepancies between research interest and availability of data were as especially important for amphibians, reptiles and fishes. Compared to studies on animal taxa, papers on plants using CS data remain rare. Although the aims and scope of papers are diverse, species conservation remained the central theme of SDM using CS data. We present examples of the use of CS and highlight recommendations to motivate further research, such as combining multiple data sources and promoting local and traditional knowledge. We hope our findings will strengthen citizen-researchers partnerships to better inform SDMs, especially for less-studied taxa and regions. Researchers stand to benefit from the large quantity of data available from CS sources to improve global predictions of species distributions.

Benefits and Drawbacks of Citizen Science to Complement Traditional Data Gathering Approaches for Medically Important Hard Ticks (Acari: Ixodidae) in the United States.

Tick-borne diseases are increasing in North America. Knowledge of which tick species and associated human pathogens are present locally can inform the public and medical community about the acarological risk for tick bites and tick-borne infections. Citizen science (also called community-based monitoring, volunteer monitoring, or participatory science) is emerging as a potential approach to complement traditional tick record data gathering where all aspects of the work is done by researchers or public health professionals. One key question is how citizen science can best be used to generate high-quality data to fill knowledge gaps that are difficult to address using traditional data gathering approaches. Citizen science is particularly useful to generate information on human-tick encounters and may also contribute to geographical tick records to help define species distributions across large areas. Previous citizen science projects have utilized three distinct tick record data gathering methods including submission of: 1) physical tick specimens for identification by professional entomologists, 2) digital images of ticks for identification by professional entomologists, and 3) data where the tick species and life stage were identified by the citizen scientist. We explore the benefits and drawbacks of citizen science, relative to the traditional scientific approach, to generate data on tick records, with special emphasis on data quality for species identification and tick encounter locations. We recognize the value of citizen science to tick research but caution that the generated information must be interpreted cautiously with data quality limitations firmly in mind to avoid misleading conclusions.

Improving big citizen science data: Moving beyond haphazard sampling.

Citizen science is mainstream: millions of people contribute data to a growing array of citizen science projects annually, forming massive datasets that will drive research for years to come. Many citizen science projects implement a "leaderboard" framework, ranking the contributions based on number of records or species, encouraging further participation. But is every data point equally "valuable?" Citizen scientists collect data with distinct spatial and temporal biases, leading to unfortunate gaps and redundancies, which create statistical and informational problems for downstream analyses. Up to this point, the haphazard structure of the data has been seen as an unfortunate but unchangeable aspect of citizen science data. However, we argue here that this issue can actually be addressed: we provide a very simple, tractable framework that could be adapted by broadscale citizen science projects to allow citizen scientists to optimize the marginal value of their efforts, increasing the overall collective knowledge.

Citizen science informs human-tick exposure in the Northeastern United States.

BACKGROUND: Tick-borne disease is the result of spillover of pathogens into the human population. Traditionally, literature has focused on characterization of tick-borne disease pathogens and ticks in their sylvatic cycles. A limited amount of research has focused on human-tick exposure in this system, especially in the Northeastern United States. Human-tick interactions are crucial to consider when assessing the risk of tick-borne disease since a tick bite is required for spillover to occur. METHODS: Citizen scientists collected ticks from the Northeastern US through a free nationwide program. Submitted ticks were identified to species, stage, and sex. Blacklegged ticks, Ixodes scapularis, were tested for the presence of Borrelia burgdorferi sensu lato (s.l.) and hard-tick relapsing fever Borrelia. Seasonality of exposure and the citizen science activity during tick exposure was recorded by the citizen scientist. A negative binomial model was fit to predict county level CDC Lyme disease cases in 2016 using citizen science Ixodes scapularis submissions, state, and county population as predictor variables. RESULTS: A total of 3740 submissions, comprising 4261 ticks, were submitted from the Northeastern US and were reported to be parasitizing humans. Of the three species submitted, blacklegged ticks were the most prevalent followed by American dog ticks and lone star ticks. Submissions peaked in May with the majority of exposure occurring during every-day activities. The most common pathogen in blacklegged ticks was B. burgdorferi s.l. followed by hard-tick relapsing fever Borrelia. Negative binomial model performance was best in New England states followed by Middle Atlantic states. CONCLUSIONS: Citizen science provides a low-cost and effective methodology for describing the seasonality and characteristics of human-tick exposure. In the Northeastern US, everyday activities were identified as a major mechanism for tick exposure, supporting the role of peri-domestic exposure in tick-borne disease. Citizen science provides a method for broad pathogen and tick surveillance, which is highly related to human disease, allowing for inferences to be made about the epidemiology of tick-borne disease.