TOPO TA is A-OK: a test of phylogenetic bias in fungal environmental clone library construction.

Abstract

TA cloning methods are widely used in analyses of environmental microbial diversity, yet the potential of TA methods to yield phylogenetically biased results has received little attention. To test for a TA bias, we constructed clone libraries of fungal amplicons spanning the ribosomal internally transcribed spacer (ITS) and partial large subunit (LSU) from 92 boreal forest soil DNA extracts using two contrasting methods: the Invitrogen TOPO-TA system and the Lucigen PCR-SMART system. The Lucigen system utilizes blunt-ended rather than TA cloning and transcription terminators to reduce biases due to toxicity of expressed inserts. We analysed 588 clone sequences from the two libraries. Species diversity estimators applied to operational taxonomical units (OTUs) were slightly higher for Invitrogen than Lucigen, but confidence intervals for accumulation curves overlapped. Abundances of OTUs were correlated between the libraries (r(2) = 0.5, P < 0.0001), but certain OTUs had contrasting abundances in the two libraries and a likelihood ratio test rejected homogeneity of the OTU counts. We constructed parsimony and Bayesian trees from aligned LSU regions, and the 'phylogenetic test' revealed that lineage representation was not significantly different between the two libraries. We conclude that characterization of this fungal community was fairly robust to cloning method and no biases due to TA cloning were found.

Full text

Page 1

TOPO TA is A-OK: a test of phylogenetic bias in fungal
environmental clone library construction
D. Lee Taylor,1* Ian C. Herriott,1James Long1and
Keith O’Neill2
1University of Alaska, Institute of Arctic Biology, 311
Irving I Building, Fairbanks, AK 99775, USA.
2Broad Institute – Massachusetts Institute for
Technology, 320 Charles St., Cambridge, MA 02141,
USA.
Summary
TA cloning methods are widely used in analyses of
environmental microbial diversity, yet the potential of
TA methods to yield phylogenetically biased results
has received little attention. To test for a TA bias, we
constructed clone libraries of fungal amplicons span-
ning the ribosomal internally transcribed spacer (ITS)
and partial large subunit (LSU) from 92 boreal forest
soil DNA extracts using two contrasting methods: the
Invitrogen TOPO-TA system and the Lucigen PCR-
SMART system. The Lucigen system utilizes blunt-
ended rather than TA cloning and transcription
terminators to reduce biases due to toxicity of
expressed inserts. We analysed 588 clone sequences
from the two libraries. Species diversity estimators
applied to operational taxonomical units (OTUs) were
slightly higher for Invitrogen than Lucigen, but confi-
dence intervals for accumulation curves overlapped.
Abundances of OTUs were correlated between the
libraries (r2= 0.5, P < 0.0001), but certain OTUs had
contrasting abundances in the two libraries and a
likelihood ratio test rejected homogeneity of the OTU
counts. We constructed parsimony and Bayesian
trees from aligned LSU regions, and the ‘phylogenetic
test’ revealed that lineage representation was not sig-
nificantly different between the two libraries. We con-
clude that characterization of this fungal community
was fairly robust to cloning method and no biases
due to TA cloning were found.
Introduction
Direct characterization of DNA from environmental
samples has revolutionized our understandings of micro-
bial evolution and ecology (e.g. Stahl et al., 1984;
Schadt et al., 2003). Efforts to enumerate the abun-
dance and diversity of microbial taxa typically involve
PCR amplification of diagnostic gene regions which
results in a heterogeneous pool of PCR products. Sense
can be made of this taxonomic slurry using fingerprinting
methods such as denaturing gradient gel electrophoresis
or terminal restriction fragment length polymorphism or
by construction of clone libraries followed by sequence
analysis. Clone library construction offers the benefit of
phylogenetically explicit data. Much effort has been
directed to characterizing and minimizing PCR artefacts
and biases (Reysenbach et al., 1992; Suzuki and Giov-
annoni, 1996; Polz and Cavanaugh, 1998), but little
attention has been paid to potential biases associated
with cloning. The most widely utilized cloning strategies
take advantage of non-template adenine additions by
standard DNA polymerases to create one-base sticky
ends complementary to the T vector. TA cloning systems
could introduce biases through several mechanisms.
First, it is known that the probability of non-template-
dependent adenine addition is related to juxtaposed
template sequence (Brownstein et al., 1996; Magnuson
et al., 1996). Thus, some taxa might be over- or under-
represented in a TA-based clone library depending on
their sequences. Second, most commercial TA vectors
include a promoter upstream of the insertion site
which allows for selection of insert-containing colonies
by variousstrategiessuch
b-galactosidase gene fragment. A transcribed insert
might interfere with growth of the transformed Escheri-
chia coli cells such that these insert sequences will not
be recovered in the resulting library. An obvious example
would be an insert encoding a toxin gene. Transcribed
rRNA inserts could also potentially interfere with ribo-
some function in the host E. coli. Third, it is known that
reaction kinetics favour insertion of shorter DNA frag-
ments over longer DNA fragments (Sambrook and
Russell, 2001), and ligation efficiency varies among
vector systems.
Our interest is in characterization of fungal communities
in soil. Here, we have constructed a library of fungal
community ribosomal PCR products with the widely used
Invitrogen TOPO-TAtopoisomerase-based cloning kit and
compared this with a library constructed from the same
samples using a blunt-ended cloning system.
asinterruption of the
Received 8 June, 2006; accepted 11 December, 2006. *For
correspondence. E-mail fflt@uaf.edu; Tel. (+1) 907 474 6982; Fax
(+1) 907 474 6967.
Environmental Microbiology (2007) 9(5), 1329–1334 doi:10.1111/j.1462-2920.2007.01253.x
© 2007 The Authors
Journal compilation © 2007 Society for Applied Microbiology and Blackwell Publishing Ltd

Page 2

Results and discussion
We amplified fungal ribosomal internally transcribed
spacer – large subunit (ITS-LSU) fragments from 92 black
spruce humic horizon soil DNA extracts and pooled the
resulting amplicons prior to cloning with (i) the Invitrogen
pCR4.0-TOPO vector and (ii) the Lucigen pcrSMART-HC
Kan vector. The Lucigen ‘low-bias’ system was chosen
because ligation to blunt-ended vectors should avoid the
potential bias related to variable non-template adenine
addition. In addition, the Lucigen vector incorporates
hairpin loops on both ends of the insertion site designed to
block insert transcription and eliminate this potential
source of bias.
Read and assembly qualities were similar in the two
libraries with 308 passing assemblies in the Lucigen
library and 312 passing assemblies in the Invitrogen
library. More short inserts were encountered in the
Lucigen library, which could be related to the end-repair
step used to remove overhanging adenines. A larger
number of non-fungal clones, many of which had the
TW13 primer at both ends, were encountered in the Invit-
rogen library and excluded. We used two contrasting
methods for chimera detection, but detected only a few
chimeras. Of the 620 assemblies, 32 were excluded (for
breakdown, see Table S1). Considering only full-length
sequences, the mean Lucigen insert size was 1342 while
the mean Invitrogen insert size was 1353, which were
not significantly different (Kruskal–Wallis rank sum test,
P = 0.96).
Fungal diversity in terms of total species richness based
on OTU (operational taxonomical unit) presence/absence
was similar in the two libraries. Sequence clustering using
Cap3 (Huang and Madan, 1999) at 90% sequence identity
yielded 148 OTUs, 55 of which occurred in both libraries.
These sequence data have been submitted to the
GenBankdatabaseunderaccessionnumbersEF433956–
EF434155. The Invitrogen library contained 110 OTUs, 46
of which were singletons.The Lucigen library contained 93
OTUs, 28 of which were singletons. Despite a slightly
higher total OTU richness in the Invitrogen library, the 95%
confidence intervals of the species accumulation curves
for the two libraries overlapped (Fig. S1), suggesting
similar levels of alpha diversity.
The fungal communities from the black spruce forest
humic soil horizon appear to be extremely diverse. Differ-
ent fungal species belonging to the same genus often
display less than 10% sequence divergence in the ITS
region (Horton, 2002; Geml et al., 2006). Hence, the OTU
groupings reported here certainly underestimate true
species diversity. Despite this conservative approach, the
species accumulation curves for neither individual library,
nor the two libraries combined, approach an asymptote
(Fig. S1). Non-parametric estimators of total OTU diver-
sity, taking into account undetected types, including Chao
1, ICE, ACE, Jackknife 1 and Jackknife 2, gave estimates
of 221–280 total OTUs in the community (Table S2), but
the 95% confidence intervals for these estimates were
large, ranging from 214 to 380, and did not stabilize
with increasing numbers of subsamples in rarefaction
analyses. In addition to OTU richness, the deep phyloge-
netic diversity we encountered is astounding: much of the
known deep diversity of fungi is represented in the large-
subunit tree, including the Chytridiomycota, Zygomycota,
Ascomycota, Basidiomycota and numerous orders and
families within them (data not shown). Ecological implica-
tions of the community structure of fungi at this site will be
discussed elsewhere.
If the two cloning methods operate equivalently, we
would expect them to have high community similarity
scores. We calculated the Jaccard and Sorenson indices,
which range from 0 when there is no species overlap to 1.0
when identical sets of species are present in two commu-
nities, using EstimateS 7.0 (Colwell and Coddington,
1994). The Sorenson index was 0.542 and Jaccard index
was 0.372; these low values would normally be interpreted
to indicate significant differences in the two communities/
libraries. However, these measures only take into account
species presence/absence, not abundance; hence, the
numerous singletons in both libraries have as much
impact on the indices as do the more abundant, shared
OTUs. Furthermore, these measures assume that all
species in each community have been detected, which is
clearly not the case for this data set, nor for most clone
library studies of microbial diversity. Chao and colleagues
(2005) have shown that such measures are consistently
biased towards underestimation of similarity in hyper-
diverse, undersampled communities, and have proposed
modified community similarity indices which take into
account abundance data, and also incorporate estimates
of the numbers of undetected taxa. These new indices
gave similarity estimates much closer to the predicted
value of 1.0 when comparing our two libraries: 0.822 for
Chao-Jaccard-Est Abundance and 0.902 for Chao-
Sorenson Est Abundance (Chao et al., 2005).
If the occurrence of an OTU in a clone library is related
to the abundance of the corresponding organism in the
community, and the two cloning methods give similar rep-
resentations of the community, we would expect a corre-
lation between the number of clones of a particular OTU
which occur in each library. Considering only OTUs which
were represented by two or more clones, a significant
positive correlation (r2= 0.5, F = 75.9447, P < 0.0001)
was found between the number clones belonging to a
particular OTU in the Lucigen library and the number of
clones of that OTU in the Invitrogen library. Rank abun-
dances of the dominant OTUs for the two libraries were
similar (Fig. 1). However, several OTUs abundant in the
1330 D. L. Taylor, I. C. Herriott, J. Long and K. O’Neill
© 2007 The Authors
Journal compilation © 2007 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 9, 1329–1334

Page 3

Lucigen library (e.g. OTUs 29, 32 and 34) were rare in the
Invitrogen library. The reverse was true in fewer instances
(OTU 28). It is conceivable that phylogenetic biases
against these OTUs operated during cloning. Upon cata-
loguing the phylogenetic affinities of all OTUs (data not
shown), we found that both OTUs 32 and 34 belong to the
Sebacinaceae (Basidiomycota). Across all OTUs, 23
clones associated with the Sebacinaceae were found in
the Lucigen library compared with five such clones in the
Invitrogen library. Further experiments would be required
to determine whether this pattern was due to sampling
error or to a systematic bias. However, it is notable that
none of the dominant types were completely missing from
either library. If abundances of OTUs between the libraries
are homogeneous, contingency table tests should give
non-significant results. However, when we compared only
OTUs with a frequency of 10 or more clones, in order to
avoid low expected cell counts, homogeneity of the
counts was rejected (r2= 0.024, P = 0.0012), using the
likelihood ratio test. The low r2suggests the frequencies
are not far different from those expected with random
sampling from a multinomial distribution. Indeed, if two of
the OTUs with contrasting frequencies are excluded, such
as 28 and 32, the likelihood ratio becomes non-significant.
In other words, the clone counts in the two libraries are
quite close to expectations for random samples from a
single population. While the use of clone counts as a
proxy for organismal abundances is tenuous due to arte-
facts, biases and sampling error inherent to PCR (Suzuki
and Giovannoni, 1996; Polz and Cavanaugh, 1998), and
perhaps also the cloning process, our results suggest that
moderately consistent species abundance results can be
obtained even for hyper-diverse fungal communities.
Of greater importance than species counts, per se, are
the phylogenetic distributions of taxa represented by the
two cloning methods. To evaluate representation of evo-
lutionary lineages in the two libraries, a LSU sequence
alignment of 203 taxa and 910 characters was con-
structed which included a representative of each OTU
encountered in each library. In Bayesian tree evaluation,
likelihoods and parameter estimates stabilized after
500 000 generations. A consensus was computed from
9500 trees after discarding the first 5000 in the burnin.
The best tree had a log likelihood of -27787.07. There
were 5093 most parsimonious trees of 6216 steps with
consistency indices of 0.255. A 50% majority rule consen-
sus tree was computed from these trees for subsequent
analyses. Overall topologies of the Bayesian and parsi-
mony consensus trees were largely congruent.
The parsimony and Bayesian trees were then exported
to MacClade 4 (Maddison and Maddison, 1989). The phy-
logenetic test, or P-test, developed by Maddison and
Slatkin (1991) and applied to microbial communities by
Martin (2002), was carried out. This test asks whether
significantly different sets of lineages occur in two
samples. The deeper and more diverse the lineage that is
unique to a sample, the more significant is the difference.
An additional character was added to the data matrix to
indicate the source library for each clone, mapped onto
the trees, and the minimum number of steps required to
explain the distribution of source library on the Bayesian
tree was determined to be 81. Mapping of the character
onto 1000 equiprobable random trees gave a range from
56 to 80 steps; hence the observed number of steps is
outside of the null distribution. For the parsimony tree,
source library required 80 steps to map onto the tree,
which was again beyond the right-hand tail of the null
distribution of 54–79 steps derived from 1000 randomiza-
tions of the parsimony tree. These results indicate that the
arrays of lineages represented in the two libraries are
more similar than would be expected by chance at
P < 0.001 (Martin, 2002). When this test has been used to
Fig. 1. Comparison of rank abundances of
dominant OTUs in the two libraries.
Test of phylogenetic bias in TA cloning1331
© 2007 The Authors
Journal compilation © 2007 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 9, 1329–1334

Page 4

compare microbial communities, the opposite result is
more often observed, namely that fewer steps are found
for the real tree than the random trees, indicating signifi-
cant dissimilarity of the communities (e.g. Martin, 2002;
Schadt et al., 2003; Stach et al., 2003; Acinas et al., 2004;
Dunfield and King, 2005; Eckburg et al., 2005; Papineau
et al., 2005). The overlapping lineage representation of
the Lucigen and Invitrogen libraries is apparent in the
nLSU tree shown in Fig. 2. Note that this analysis takes
phylogenetic scale into account, but is based on
presence/absence, and hence does not take OTU abun-
dance into account.
Based on the potential filtering effects of insert size,
adenine addition and expression of toxic inserts, we pre-
dicted that compared with an unbiased method the Invit-
rogen TOPO-TA cloning method would reveal less fungal
diversity and might miss entire lineages of fungi. However,
our results suggest that the contrasting cloning strategies
behave similarly under our conditions. Despite our
emphasis on potential biases of TA cloning methods, it is
perhaps not surprising that little difference was found.
Non-template adenine addition, a requisite of TA cloning,
is known to depend on neighbouring sequences, which
could bias adenine addition towards or against particular
taxa. However, this effect would only be important if it
extended beyond the 18–22 base pair (bp) primers used
for the PCR, which has not been shown in previous
studies (Brownstein et al., 1996). Size bias, though clearly
demonstrated and known to vary among cloning vectors,
is also unlikely to exert a major bias in the size range of
our inserts (1100–2000 bp). Of greatest concern, a priori,
was the possible filtering effect of ribosomal inserts
expressed from the TAvector that might interfere with host
E. coli. We found no strong evidence for such a bias, as
both libraries revealed similar and largely overlapping
phylogenetic diversity (Fig. 2). It is possible that our des-
ignation of OTUs based on pairwise divergence values
Fig. 2. Bayesian tree of the ribosomal LSU region for
representatives of each OTU from each library. There is similar
phylogenetic representation across the two libraries. The number of
‘evolutionary transitions’ from Invitrogen to Lucigen has been traced
on the tree. Black terminal branches and clone names beginning in
‘T’ are from the Invitrogen library. White terminal branches and
clone names beginning with letters other than ‘T’ are from the
Lucigen library. To prevent low-quality sequences from confounding
the phylogenetic analyses, all consensus base calls with quality
scores below 20 were converted to Ns. The ITS regions of
divergent OTUs could not be aligned, so we focused only on the
LSU region for these analyses. Large subunit regions representing
each OTU found in each library were aligned with CLUSTALX
(Thompson et al., 1997). The resulting alignment was improved by
eye. The GTR model with a portion of invariant sites and gamma
distributed rate variation across other sites, was then utilized for
tree searches with a heating parameter of 0.3 and eight chains run
for 2 000 000 generations in MrBayes 3.1 (Huelsenbeck and
Ronquist, 2001). Burnin was set to 500 000 generations based on
output from an initial run.
1332 D. L. Taylor, I. C. Herriott, J. Long and K. O’Neill
© 2007 The Authors
Journal compilation © 2007 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 9, 1329–1334

Page 5

contributed error to the comparison of libraries. While a
full phylogenetic analysis of all clones would have been a
preferable approach to OTU discrimination, the ITS region
cannot be aligned across such diverse fungi. However, we
did construct trees of all clones using only the nLSU
portion of the sequences, and found that the OTUs cor-
responded very well with clades in the LSU tree. The
differences in abundances of a few of the dominant OTUs
(Fig. 1) could be due to stochastic events during PCR,
ligation and colony picking steps or to a systematic bias
such as insert toxicity. Without repeating library construc-
tion and sequencing from numerous PCR replicates,
these potential sources of variance cannot be quantified.
Such a study would be cost-prohibitive if applied to a
hyper-diverse community due to the massive sequencing
effort required. The extreme diversity and concomitant
lack of replication of clone library construction is a weak-
ness of the experimental design used here. However, use
of a less diverse community would reduce the phyloge-
netic breadth and therefore generality of the comparison.
Even without replication of clone library construction, our
analysis clearly shows that the contrasting cloning
methods provide similar pictures of fungal phylogenetic
diversity and community composition at our study site.
In conclusion, our study has demonstrated largely
congruent pictures of OTU presence/absence, OTU
abundance and deeper phylogenetic diversity using two
contrasting cloning methods. Our results should be
welcome news to microbial ecologists attempting to
document diversity using TA cloning methods.
Acknowledgements
We thank Roger Ruess and Jack McFarland for assistance
with soil collection and Tom Marr and Shawn Huston for
computational help. Supported by NSF EF 0333308 and the
Bonanza Creek LTER (Long-Term Ecological Research)
program (funded jointly by NSF grant DEB-0423442 and
USDA Forest Service, Pacific North-west Research Station
Grant PNW01-JV11261952-231).
References
Acinas, S.G., Klepac-Ceraj, V., Hunt, D.E., Pharino, C.,
Ceraj, I., Distel, D.L., and Polz, M.F. (2004) Fine-scale
phylogenetic architecture
community. Nature 430: 551–554.
Brownstein, M.J., Carpten, J.D., and Smith, J.R. (1996)
Modulation of non-templated nucleotide addition by taq
DNA polymerase: primer modifications that facilitate
genotyping. Biotechniques
1010.
Chao, A., Chazdon, R.L., Colwell, R.K., and Shen, T.J.
(2005) Anew statistical approach for assessing similarity of
species composition with incidence and abundance data.
Ecol Lett 8: 148–159.
ofa complexbacterial
20:
1004–1006, 1008–
Colwell, R.K., and Coddington, J.A. (1994) Estimating terres-
trial biodiversity through extrapolation. Philos Trans R Soc
Lond B Biol Sci 345: 101–118.
Dunfield, K.E., and King, G.M. (2005) Analysis of the distri-
bution and diversity in recent Hawaiian volcanic deposits of
a putative carbon monoxide dehydrogenase large subunit
gene. Environ Microbiol 7: 1405–1412.
Eckburg, P.B., Bik, E.M., Bernstein, C.N., Purdom, E., Deth-
lefsen, L., Sargent, M., et al. (2005) Diversity of the human
intestinal microbial flora. Science 308: 1635–1638.
Geml, J., Laursen, G.A., O’Neill, K., Nusbaum, H.C., and
Taylor, D.L. (2006) Beringian origins and cryptic speciation
events in the fly agaric (Amanita muscaria). Mol Ecol 15:
225–239.
Horton, T.R. (2002) Molecular approaches to ectomycorrhizal
diversity studies: variation in ITS at a local scale. Plant Soil
244: 29–39.
Huang, X.Q., and Madan, A. (1999) Cap3: a DNA sequence
assembly program. Genome Res 9: 868–877.
Huelsenbeck, J.P., and Ronquist, F. (2001) MRBAYES:
bayesian inference of phylogenetic trees. Bioinformatics
(Oxford) 17: 754–755.
Maddison, W.P., and Maddison, D.R. (1989) Interactive
analysis of phylogeny and character evolution using the
computer program MacClade. Folia Primatol 53: 190–202.
Maddison, W.P., and Slatkin, M. (1991) Null models for the
number of evolutionary steps in a character on a phyloge-
netic tree. Evolution 45: 1184–1197.
Magnuson, V.L., Ally, D.S., Nylund, S.J., Karanjawala, Z.E.,
Rayman,J.B.,Knapp,J.I.,
nucleotide-determined non-templated addition of adenine
by Taq DNApolymerase: implications for PCR-based geno-
typing and cloning. Biotechniques 21: 700–709.
Martin, A.P. (2002) Phylogenetic approaches for describing
and comparing the diversity of microbial communities. Appl
Environ Microbiol 68: 3673–3682.
Papineau, D., Walker, J.J., Mojzsis, S.J., and Pace, N.R.
(2005) Composition and structure of microbial communities
from stromatolites of Hamelin Pool in Shark Bay, Western
Australia. Appl Environ Microbiol 71: 4822–4832.
Polz, M.F., and Cavanaugh, C.M. (1998) Bias in template-to-
product ratios in multitemplate PCR. Appl Environ Micro-
biol 64: 3724–3730.
Reysenbach, A.L., Giver, L.J., Wickham, G.S., and Pace,
N.R. (1992) Differential amplification of ribosomal-RNA
genes by polymerase chain-reaction. Appl Environ Micro-
biol 58: 3417–3418.
Sambrook, J., and Russell, D.W. (2001) Molecular Cloning: A
Laboratory Manual. Cold Spring Harbor, NY, USA: Cold
Spring Harbor Laboratory Press.
Schadt, C.W., Martin, A.P., Lipson, D.A., and Schmidt, S.K.
(2003) Seasonal dynamics of previously unknown fungal
lineages in tundra soils. Science 301: 1359–1361.
Stach, J.E.M., Maldonado, L.A., Masson, D.G., Ward, A.C.,
Goodfellow,M.,andBull,
approaches for estimating Actinobacterial diversity in
marine sediments. Appl Environ Microbiol 69: 6189–6200.
Stahl, D., Lane, D., Olsen, G., and Pace, N. (1984) Analysis
of hydrothermal vent-associated symbionts by ribosomal
RNA sequences. Science 224: 409–411.
Suzuki, M.T., and Giovannoni, S.J. (1996) Bias caused by
et al. (1996) Substrate
A.T. (2003)Statistical
Test of phylogenetic bias in TA cloning1333
© 2007 The Authors
Journal compilation © 2007 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 9, 1329–1334

Page 6

template annealing in the amplification of mixtures of 16S
rRNA genes by PCR. Appl Environ Microbiol 62: 625–630.
Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F.,
and Higgins, D.G. (1997) The CLUSTAL-X windows inter-
face: flexible strategies for multiple sequence alignment
aided by quality analysis tools. Nucleic Acids Res 25:
4876–4882.
Supplementary material
The following supplementary material is available for this
article online:
Fig. S1. Species accumulation curves. The numbers of
OTUs detected relative to the number of clones sequenced
are statistically equivalent, showing that the two methods
detect similar levels of diversity. Error bars represent 95%
confidence intervals for numbers of species (Mao Tao) gen-
erated by EstimateS 7.0 (Colwell and Coddington, 1994). To
assess the phylogenetic representation across the two librar-
ies, sequences were clustered into OTUs using the program
Cap3 (Huang and Madan, 1999). Quality scores exported
from Aligner were used in Cap3 to clip bases with scores
below 10, and to consider only bases with combined qualities
above 40 in determining mismatches between sequences.All
parameters were set to lenient values favouring assembly of
pairs of sequences except the minimum per cent identity in
the overlapping region, which was set to 90%, and the
maximum overhang per cent length, which was set to 20%.
To create species (OTU) accumulation curves for each
library, the observed individuals were randomly distributed to
create 50 mock samples for each library, with 50 repetitions of
this randomization process.
Table S1. Summary of clones used for analyses.
Table S2. Diversity and similarity statistics for the Invitrogen
and Lucigen clone libraries.
This material is available as part of the online article from
http://www.blackwell-synergy.com
1334D. L. Taylor, I. C. Herriott, J. Long and K. O’Neill
© 2007 The Authors
Journal compilation © 2007 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 9, 1329–1334