An ecoinformatics tool for microbial community studies: Supervised classification of Amplicon Length Heterogeneity (ALH) profiles of 16S rRNA

Abstract

Support vector machines (SVM) and K-nearest neighbors (KNN) are two computational machine learning tools that perform supervised classification. This paper presents a novel application of such supervised analytical tools for microbial community profiling and to distinguish patterning among ecosystems. Amplicon length heterogeneity (ALH) profiles from several hypervariable regions of 16S rRNA gene of eubacterial communities from Idaho agricultural soil samples and from Chesapeake Bay marsh sediments were separately analyzed. The profiles from all available hypervariable regions were concatenated to obtain a combined profile, which was then provided to the SVM and KNN classifiers. Each profile was labeled with information about the location or time of its sampling. We hypothesized that after a learning phase using feature vectors from labeled ALH profiles, both these classifiers would have the capacity to predict the labels of previously unseen samples. The resulting classifiers were able to predict the labels of the Idaho soil samples with high accuracy. The classifiers were less accurate for the classification of the Chesapeake Bay sediments suggesting greater similarity within the Bay's microbial community patterns in the sampled sites. The profiles obtained from the VI +V2 region were more informative than that obtained from any other single region. However, combining them with profiles from the V1 region (with or without the profiles from the V3 region) resulted in the most accurate classification of the samples. The addition

Bacterial community structures are unique and resilient in full-scale bioenergy systems

“…Due to the size and variability of the OTU table, we identified predictive OTUs using sparse supervised learning, which has been successfully applied to classification of microarray data (19) and 16S amplicon length heterogeneity profiles (20). We chose location of the facility as the class attribute to be predicted, because the community phylogenetic structure was observed to cluster most closely on the basis of facility location, and because facilities were also unique and consistent in terms of performance data (Fig.…”

Assessing Microbial Community Diversity Using Amplicon Length Heterogeneity Polymerase Chain Reaction

“…In two recent studies by our group, LH-PCR was used to query which hypervariable domain or combination of 16S rRNA gene domains was the best molecular marker Yang et al, 2006). In the first study, data from Idaho natural sagebrush and irrigated moldboard plowed sites were used to compare univariate and multivariate analyses.…”

Amplicon length heterogeneity (ALH)-PCR generated bacterial community profiling : a novel application for the forensic examination of soil

“…The "pruned" data was then pasted into rows combining the data from V1, V3 and Vi + V2 regions (113). All data in each row represented PCR's from a single soil extraction.…”

“…All data in each row was normalized (individual peak fluorescence divided by total fluorescence of all peaks in row) so that each fragment length now had a relative abundance unit attached. At this point another "pruning" was performed removing any peak representing less than 1 % (0.01) of the total profile (113). The resulting microbial community profile is our final data output including fragment length in bp and relative abundance of each peak in the profile from each soil replicate.…”

“…In recent years, DNA-based profiling of microbial communities has become a rapid and inexpensive process (107), requiring only two or three days to process multiple samples (~ 30), and has become quite routine, requiring relatively little training. The profiles derived from these methods can be used to distinguish between soils of interest (40,76,88,113).…”