Dataset Information

Predicting microbial traits with phylogenies.

ABSTRACT: Phylogeny reflects genetic and phenotypic traits in Bacteria and Archaea. The phylogenetic conservatism of microbial traits has prompted the application of phylogeny-based algorithms to predict unknown trait values of extant taxa based on the traits of their evolutionary relatives to estimate, for instance, rRNA gene copy numbers, gene contents or tolerance to abiotic conditions. Unlike the 'macrobial' world, microbial ecologists face scenarios potentially compromising the accuracy of trait reconstruction methods, as, for example, extremely large phylogenies and limited information on the traits of interest. We review 990 bacterial and archaeal traits from the literature and support that phylogenetic trait conservatism is widespread through the tree of life, while revealing that it is generally weak for ecologically relevant phenotypic traits and high for genetically complex traits. We then perform a simulation exercise to assess the accuracy of phylogeny-based trait predictions in common scenarios faced by microbial ecologists. Our simulations show that ca. 60% of the variation in phylogeny-based trait predictions depends on the magnitude of the trait conservatism, the number of species in the tree, the proportion of species with unknown trait values and the mean distance in the tree to the nearest neighbour with a known trait value. Results are similar for both binary and continuous traits. We discuss these results under the light of the reviewed traits and provide recommendations for the use of phylogeny-based trait predictions for microbial ecologists.

SUBMITTER: Goberna M

PROVIDER: S-EPMC4796935 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:The assumption that traits and phylogenies can be used as proxies of species niche has faced criticisms. Evidence suggested that phylogenic relatedness is a weak proxy of trait similarity. Moreover, different processes can select different traits, giving opposing signals in null model analyses. To circumvent these criticisms, we separated traits of stream insects based on the concept of ? and ? niches, which should give clues about assembling pressures expected to act independently of each other. We investigated the congruence between the phylogenetic structure and trait structure of communities using all available traits and all possible combinations of traits (4095 combinations). To account for hierarchical assembling processes, we analyzed patterns on two spatial scales with three pools of genera. Beta niche traits selected a priori - i.e., traits related to environmental variation (e.g., respiration type) - were consistently clustered on the smaller scale, suggesting environmental filtering, while ? niche traits - i.e., traits related to resource use (e.g., trophic position) - did not display the expected overdispersion, suggesting a weak role of competition. Using all traits together provided random patterns and the analysis of all possible combinations of traits provided scenarios ranging from strong clustering to overdispersion. Communities were phylogenetically overdispersed, a pattern previously interpreted as phylogenetic limiting similarity. However, our results likely reflect the co-occurrence of ancient clades due to the stability of stream habitats along the evolutionary scale. We advise ecologists to avoid using combinations of all available traits but rather carefully traits based on the objective under consideration. Both trait and phylogenetic approaches should be kept in the ecologist toolbox, but phylogenetic distances should not be used as proxies of traits differences. Although the phylogenetic structure revealed processes operating at the evolutionary scale, only specific traits explained local processes operating in our communities.

Project description:Fusarium head blight (FHB) incited by Fusarium graminearum Schwabe is a devastating disease of barley and other cereal crops worldwide. Fusarium head blight is associated with trichothecene mycotoxins such as deoxynivalenol (DON), where contaminated grains are unfit for malting or animal feed industries. While genetically resistant cultivars offer the best economic and environmentally responsible means to mitigate disease, parent lines with adequate resistance are limited in barley. Resistancebreeding based upon quantitative genetic gains has been slow to date, due to intensive labour requirements of disease nurseries. The development of high throughput genome-wide molecular markers, allow application in genomic prediction models. A diverse genomic panel consisting of 400 two-row spring barley lines was assembled to focus on Canadian barley breeding programs. The panel was evaluated for FHB and DON content in three environments and over two years. Moreover, it was genotyped using an Illumina Infinium HTS iSelect custom beadchip array of single nucleotide polymorphic molecular markers (50K SNP), where over 23K molecular markers were polymorphic. Genomic prediction has been successfully demonstrated for reducing FHB and DON content in cereals using various statistically-based models of different underlying assumptions. Herein, we have studied an alternative method basedon machine learning and compare it with a statistical approach. Two encoding techniques were utilized (categorical or Hardy-Weinberg frequencies), followed by selecting essential genomic markers for phenotype prediction. Subsequently, we applied a transformer-based deep learning algorithm to predict FHB and DON. Apart from the transformer method, we also implemented a Residual Fully Connected Neural Network (RFCNN). Pearson correlation coefficients were calculated to compare true vs. predicted outputs. Under most model scenarios, the use of all markers vs. selected markers marginally improved prediction performance except for RFCNN method for FHB (27.6%). Hardy-Weinberg encoding generally improved correlation for FHB (6.9%) and DON (9.6%) for transformer. This study suggests the potential of the transformer based method for genomic prediction of complex traits such as FHB or DON, having performed better or equally compared with existing machine learning and statistical method. To genomic prediction in barley for Fusarium head blight and deoxynivalenol content using a custom Illumina Infinium array (BarleySNP50-JHI) (www.illumina.com). Sample types included leaves from 400 barley genotypes mostly of Canadian origin. This series includes 400 genotypes assayed on an Illumina infinium HTS platform 50K BeadChip.

Dataset Information

Predicting microbial traits with phylogenies.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets