mebioda

Metabarcoding

General workflow of metabarcoding assays

• sequence pre-processing, e.g. de-replication, filtering low-quality and overly short reads, trimming low-quality ends, merging pairs. Using generic HTS tools.
• demultiplexing, e.g. to split by sampling locations and/or times. Done, for example, with QIIME, OBITools or mothur
• clustering using tools such as CD-HIT, UCLUST, or OCTUPUS
• outlier detection, e.g. chimeric sequences or singletons
• taxonomic assignment, e.g. by BLAST searches against reference databases, or with usearch or vsearch
• phylogenetic analysis, e.g. phylogenetic placement, computation of diversity metrics
• comparing treatments, e.g. by rarefaction of OTU tables

Species identification of gut contents of permafrost grazers

• ancient DNA sequencing of the gut contents of permafrost grazers
• two chloroplast markers (rbcL and trnL-trnF) amplified with forward and reverse primers
• findings corroborated with morphological analysis of macroremains and pollen

Analysis workflow:

1. demultiplex on IonTorrent adaptors; Phred quality (Q20) and length (>=100bp) filter
2. cluster reads with CD-HIT
3. BLAST against NCBI nr

Mid-Holocene horse

B Gravendeel, A Protopopov, I Bull, E Duijm, F Gill, A Nieman, N Rudaya, A N Tikhonov, S Trofimova, GBA van Reenen, R Vos, S Zhilich & B van Geel. 2014. Multiproxy study of the last meal of a mid-Holocene Oyogos Yar horse, Sakha Republic, Russia. The Holocene 24(10): 1288-1296 doi:10.1177/0959683614540953

• c. 5,400 years ago, Oyogos Yar, Russia
• Pollen grains and the aDNA record give information about taxa that occurred in the landscape.
• The combined data point to an open landscape of a coastal tundra dominated by graminoids (Poaceae, Cyperaceae) with a limited amount of Birch and Alder.

Early Holocene Yakutian bison

B van Geel, A Protopopov, I Bull, E Duijm, F Gill, Y Lammers, A Nieman, N Rudaya, S Trofimova, A N Tikhonov, R Vos, S Zhilich, B Gravendeel. 2014. Multiproxy diet analysis of the last meal of an early Holocene Yakutian bison. Journal of Quaternary Science 29(3): 261-268 doi:10.1002/jqs.2698

• c. 10,500 years ago, Chuckchalakh lake, Russia
• Remains of shrubs (Alnus, Betula, Salix) and Poaceae indicate that the animal probably lived in a landscape of predominantly dry soils, intermixed with wetlands containing herbaceous plant species, as indicated by remains of Comarum palustre, Caltha palustris, Eriophorum, Sparganium, Menyanthes trifoliata and Utricularia.
• All recorded taxa still occur in the present-day Yakutian tundra vegetation.

Joining CITES listing with species detection in organic mixtures

Y Lammers, T Peelen, R A Vos & B Gravendeel. 2014. The HTS barcode checker pipeline, a tool for automated detection of illegally traded species from high-throughput sequencing data. BMC Bioinformatics 15:44 doi:10.1186/1471-2105-15-44

First, species from the CITES appendices are joined with the species in the NCBI taxonomy using the GlobalNames taxonomic name resolution service. A simple request to this service would be:

curl -o globalnames.json http://resolver.globalnames.org/name_resolvers.json?names=Homo+sapiens

Which results in a large JSON file. In a script, you might process such data as follows:

# doing a single request to the globalnames service
# usage: tnrs.py 'Homo sapiens'
import requests, sys
try:
	url = 'http://resolver.globalnames.org/name_resolvers.json'
	response = requests.get(url, params={'names':sys.argv[1]}, allow_redirects=True)
	json = response.json()
except:
	json = []

if 'results' in json['data'][0].keys():
	if 'name_string' in json['data'][0]['results'][0].keys():
		for data_dict in json['data']:
			for results_dict in data_dict['results']:
				if results_dict['data_source_title'] == 'NCBI':
					print( results_dict['taxon_id'] )					

Using this logic, a file-based database is populated:

Querying the CITES-annotated reference database

Comparing treatments: metabarcoding the Deepwater Horizon oil spill

HM Bik, KM Halanych, J Sharma & WK Thomas. 2012. Dramatic Shifts in Benthic Microbial Eukaryote Communities following the Deepwater Horizon Oil Spill. PLoS ONE 7(6): e38550 doi:10.1371/journal.pone.0038550

Deepwater Horizon sampling design

A study using 454 data processed with the QIIME pipeline. With these data the assumption was that the reads are structured according to the following primer and amplicon construct:

In this case with data with the following experimental design:

• sampled over two points in time (pre- and post-spill);
• in 7 localities (Bayfront Park, Shellfish Lab, Ryan Ct, Cadillac Ave, Dauphin Bay, Belleair Blvd, Grand Isle), mostly in the vicinity of Mobile, Alabama
• sequencing two markers (regions of the 18 S gene) with two primers (respectively F04/R22, NF1/18Sr2b)

Oil spill impact: dramatic shifts in benthic microbial eukaryote communities

Accordingly, the reads were demultiplexed following this complex mapping. The reads were then clustered with UCLUST and denoised. Finally, taxonomic identification of each cluster was performed using MegaBLAST, resulting in a sample by taxon table alternatively visualized as follows:

Phylogenetic diversity

To quantify the turnover between sites and treatments, it is useful to compute metrics of phylogenetic β diversity, such as UniFrac.

Squares, triangles, and circles denote sequences derived from different communities. Branches attached to nodes are colored black if they are unique to a particular environment and gray if they are shared.

• A Tree representing phylogenetically similar communities, where a significant fraction of the branch length in the tree is shared (gray).
• B Tree representing two communities that are maximally different so that 100% of the branch length is unique to either the circle or square environment.

• C Using the UniFrac metric to determine if the circle and square communities are significantly different. For n replicates (r), the environment assignments of the sequences were randomized, and the fraction of unique (black) branch lengths was calculated. The reported P value is the fraction of random trees that have at least as much unique branch length as the true tree (arrow). If this P value is below a defined threshold, the samples are considered to be significantly different.

• D The UniFrac metric can be calculated for all pairwise combinations of environments in a tree to make a distance matrix. This matrix can be used with standard multivariate statistical techniques such as UPGMA and principal coordinate analysis to compare the biotas in the environments.

Constructing a large metabarcoding phylogeny

• A common approach is to align ribosomal RNA against the SILVA reference database (in this case using PyNAST) and infer a tree (e.g. with FastTree)
• This tree then becomes the input tree for the Unifrac distance calculations (β diversity), resulting in a distance matrix
• From the distance matrix, a tree can clustered that, in this case, shows the similarity among the post-spill sites

Principal Coordinate Analysis

• Because the Unifrac distance matrix is multidimensional, methods to reduce dimensionality (e.g. to 3D) such as PCoA, to explore and to visualize similarities or dissimilarities of data. It starts with a similarity matrix or dissimilarity matrix (= distance matrix) and assigns for each item a location in a low-dimensional space, e.g. as a 3D graphics
• The 2D visualization broadly shows the same as the environment clustering: post-spill sites are similar to one another.

The authors claim that the pre-spill sites were more diverse. This is not very obvious from the 2D plot, but perhaps it is clearer in an interactive 3D view.

The supplementary data with the paper have PCoA results in *.kin files next to folders that have a king.jar file in them. Use this to view some of the *.kin files. Were the pre-spill sites more diverse?

This site is open source. Improve this page.