Dataset Information

Selection of optimal proxy locations for temperature field reconstructions using evolutionary algorithms.

ABSTRACT: In the Era of exponential data generation, increasing the number of paleoclimate records to improve climate field reconstructions might not always be the best strategy. By using pseudo-proxies from different model ensembles, we show how biologically-inspired artificial intelligence can be coupled with different reconstruction methods to minimize the spatial bias induced by the non-homogeneous distribution of available proxies. The results indicate that small subsets of records situated over representative locations can outperform the reconstruction skill of the full proxy network, even in more realistic pseudo-proxy experiments and observational datasets. These locations highlight the importance of high-latitude regions and major teleconnection areas to reconstruct annual global temperature fields and their responses to external forcings and internal variability. However, low frequency temperature variations such as the transition between the Medieval Climate Anomaly and the Little Ice Age are better resolved by records situated at lower latitudes. According to our idealized experiments a careful selection of proxy locations should be performed depending on the targeted time scale of the reconstructed field.

SUBMITTER: Jaume-Santero F

PROVIDER: S-EPMC7221084 | biostudies-literature | 2020 May

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Selection of optimal proxy locations for temperature field reconstructions using evolutionary algorithms.

Jaume-Santero Fernando F Barriopedro David D García-Herrera Ricardo R Calvo Natalia N Salcedo-Sanz Sancho S

Scientific reports 20200513 1

In the Era of exponential data generation, increasing the number of paleoclimate records to improve climate field reconstructions might not always be the best strategy. By using pseudo-proxies from different model ensembles, we show how biologically-inspired artificial intelligence can be coupled with different reconstruction methods to minimize the spatial bias induced by the non-homogeneous distribution of available proxies. The results indicate that small subsets of records situated over repr ...[more]

PMID: 32404961

Similar Datasets

Project description:BackgroundStudies on genome-wide associations help to determine the cause of many genetic diseases. Genome-wide associations typically focus on associations between single-nucleotide polymorphisms (SNPs). Genotyping every SNP in a chromosomal region for identifying genetic variation is computationally very expensive. A representative subset of SNPs, called tag SNPs, can be used to identify genetic variation. Small tag SNPs save the computation time of genotyping platform, however, there could be missing data or genotyping errors in small tag SNPs. This study aims to solve Tag SNPs selection problem using many-objective evolutionary algorithms.MethodsTag SNPs selection can be viewed as an optimization problem with some trade-offs between objectives, e.g. minimizing the number of tag SNPs and maximizing tolerance for missing data. In this study, the tag SNPs selection problem is formulated as a many-objective problem. Nondominated Sorting based Genetic Algorithm (NSGA-III), and Multi-Objective Evolutionary Algorithm based on Decomposition (MOEA/D), which are Many-Objective evolutionary algorithms, have been applied and investigated for optimal tag SNPs selection. This study also investigates different initialization methods like greedy and random initialization. optimization.ResultsThe evaluation measures used for comparing results for different algorithms are Hypervolume, Range, SumMin, MinSum, Tolerance rate, and Average Hamming distance. Overall MOEA/D algorithm gives superior results as compared to other algorithms in most cases. NSGA-III outperforms NSGA-II and other compared algorithms on maximum tolerance rate, and SPEA2 outperforms all algorithms on average hamming distance.ConclusionExperimental results show that the performance of our proposed many-objective algorithms is much superior as compared to the results of existing methods. The outcomes show the advantages of greedy initialization over random initialization using NSGA-III, SPEA2, and MOEA/D to solve the tag SNPs selection as many-objective optimization problem.

Project description:Evaluation of surgical skills during minimally invasive surgeries is needed when recruiting new surgeons. Although surgeons' differentiation by skill level is highly complex, performance in specific clinical tasks such as pegboard transfer and knot tying could be determined using wearable EMG and accelerometer sensors. A wireless wearable platform has made it feasible to collect movement and muscle activation signals for quick skill evaluation during surgical tasks. However, it is challenging since the placement of multiple wireless wearable sensors may interfere with their performance in the assessment. This study utilizes machine learning techniques to identify optimal muscles and features critical for accurate skill evaluation. This study enrolled a total of twenty-six surgeons of different skill levels: novice (n = 11), intermediaries (n = 12), and experts (n = 3). Twelve wireless wearable sensors consisting of surface EMGs and accelerometers were placed bilaterally on bicep brachii, tricep brachii, anterior deltoid, flexor carpi ulnaris (FCU), extensor carpi ulnaris (ECU), and thenar eminence (TE) muscles to assess muscle activations and movement variability profiles. We found features related to movement complexity such as approximate entropy, sample entropy, and multiscale entropy played a critical role in skill level identification. We found that skill level was classified with highest accuracy by i) ECU for Random Forest Classifier (RFC), ii) deltoid for Support Vector Machines (SVM) and iii) biceps for Naïve Bayes Classifier with classification accuracies 61%, 57% and 47%. We found RFC classifier performed best with highest classification accuracy when muscles are combined i) ECU and deltoid (58%), ii) ECU and biceps (53%), and iii) ECU, biceps and deltoid (52%). Our findings suggest that quick surgical skill evaluation is possible using wearables sensors, and features from ECU, deltoid, and biceps muscles contribute an important role in surgical skill evaluation.

Project description:BackgroundArtificial selection on quantitative traits using breeding values and selection indices in commercial livestock breeding populations causes changes in allele frequency over time at hundreds or thousands of causal loci and the surrounding genomic regions. In population genetics, this type of selection is called polygenic selection. Researchers and managers of pig breeding programs are motivated to understand the genetic basis of phenotypic diversity across genetic lines, breeds, and populations using selection mapping analyses. Here, we applied generation proxy selection mapping (GPSM), a genome-wide association analysis of single nucleotide polymorphism (SNP) genotypes (38,294-46,458 markers) of birth date, in four pig populations (15,457, 15,772, 16,595 and 8447 pigs per population) to identify loci responding to artificial selection over a period of five to ten years. Gene-drop simulation analyses were conducted to provide context for the GPSM results. Selected loci within and across each population of pigs were compared in the context of swine breeding objectives.ResultsThe GPSM identified 49 to 854 loci as under selection (Q-values less than 0.10) across 15 subsets of pigs based on combinations of populations. The number of significant associations increased when data were pooled across populations. In addition, several significant associations were identified in more than one population. These results indicate concurrent selection objectives, similar genetic architectures, and shared causal variants responding to selection across these pig populations. Negligible error rates (less than or equal to 0.02%) of false-positive associations were found when testing GPSM on gene-drop simulated genotypes, suggesting that GPSM distinguishes selection from random genetic drift in actual pig populations.ConclusionsThis work confirms the efficacy and the negligible error rates of the GPSM method in detecting selected loci in commercial pig populations. Our results suggest shared selection objectives and genetic architectures across swine populations. The identified polygenic selection highlights loci that are important to swine production.

Project description:Small mammal species play an important role influencing vegetation primary productivity and plant species composition, seed dispersal, soil structure, and as predator and/or prey species. Species which experience population dynamics cycles can, at high population phases, heavily impact agricultural sectors and promote rodent-borne disease transmission. To better understand the drivers behind small mammal distributions and abundances, and how these differ for individual species, it is necessary to characterise landscape variables important for the life cycles of the species in question. In this study, a suite of Earth observation derived metrics quantifying landscape characteristics and dynamics, and in-situ small mammal trapline and transect survey data, are used to generate random forest species distribution models for nine small mammal species for study sites in Narati, China and Sary Mogul, Kyrgyzstan. These species distribution models identify the important landscape proxy variables driving species abundance and distributions, in turn identifying the optimal conditions for each species. The observed relationships differed between species, with the number of landscape proxy variables identified as important for each species ranging from 3 for Microtus gregalis at Sary Mogul, to 26 for Ellobius tancrei at Narati. Results indicate that grasslands were predicted to hold higher abundances of Microtus obscurus, E. tancrei and Marmota baibacina, forest areas hold higher abundances of Myodes centralis and Sorex asper, with mixed forest-grassland boundary areas and areas close to watercourses predicted to hold higher abundances of Apodemus uralensis and Sicista tianshanica. Localised variability in vegetation and wetness conditions, as well as presence of certain habitat types, are also shown to influence these small mammal species abundances. Predictive application of the Random Forest (RF) models identified spatial hot-spots of high abundance, with model validation producing R2 values between 0.670 for M. gregalis transect data at Sary Mogul to 0.939 for E. tancrei transect data at Narati. This enhances previous work whereby optimal habitat was defined simply as presence of a given land cover type, and instead defines optimal habitat via a combination of important landscape dynamic variables, moving from a human-defined to species-defined perspective of optimal habitat. The species distribution models demonstrate differing distributions and abundances of host species across the study areas, utilising the strengths of Earth observation data to improve our understanding of landscape and ecological linkages to small mammal distributions and abundances.

Dataset Information

Selection of optimal proxy locations for temperature field reconstructions using evolutionary algorithms.

Publications

Selection of optimal proxy locations for temperature field reconstructions using evolutionary algorithms.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets