[2014.08.25] Albertsen ISME15 CAMI: Why metgenomics is broken

22
Why metagenomics is broken Mads Albertsen, PhD student Søren M. Karst, Rasmus H. Kirekegaard & Per H. Nielsen CENTER FOR MICROBIAL COMMUNITIES MI -08-2014 ME15, Soul, South Korea @MadsAlbertsen85 slides: goo.gl/xxxxx

description

Why metgenomics is broken

Transcript of [2014.08.25] Albertsen ISME15 CAMI: Why metgenomics is broken

Page 1: [2014.08.25] Albertsen ISME15 CAMI: Why metgenomics is broken

Why metagenomics

is broken

Mads Albertsen, PhD studentSøren M. Karst, Rasmus H. Kirekegaard & Per H. Nielsen

CENTER FOR MICROBIAL COMMUNITIES

CAMI25-08-2014ISME15, Soul, South Korea

@MadsAlbertsen85

slides: goo.gl/xxxxx

Page 2: [2014.08.25] Albertsen ISME15 CAMI: Why metgenomics is broken

CENTER FOR MICROBIAL COMMUNITIES | AALBORG UNIVERSITYwww.cami-challenge.org

Critical Assesment of Metagenome Interpretation

Showcase potentials and limitations

Page 3: [2014.08.25] Albertsen ISME15 CAMI: Why metgenomics is broken

CENTER FOR MICROBIAL COMMUNITIES | AALBORG UNIVERSITYwww.cami-challenge.org

Critical Assesment of Metagenome Interpretation

Showcase potentials and limitations

The implication of lack of reference genomes on taxonomic classification

The impact of micro-diversity onassembly and

genome completeness metrics

More visualisations, please.

Page 4: [2014.08.25] Albertsen ISME15 CAMI: Why metgenomics is broken

CENTER FOR MICROBIAL COMMUNITIES | AALBORG UNIVERSITY

Sequence

Assemble

Compare to database(BLAST vs. NCBI RefSeq)

Taxonomic Classification(MEGAN)

Page 5: [2014.08.25] Albertsen ISME15 CAMI: Why metgenomics is broken

CENTER FOR MICROBIAL COMMUNITIES | AALBORG UNIVERSITY

Page 6: [2014.08.25] Albertsen ISME15 CAMI: Why metgenomics is broken

CENTER FOR MICROBIAL COMMUNITIES | AALBORG UNIVERSITY

The simplest metagenome in

the world

Accumulibacter (Class Betaproteobacteria) Nitrospira (Class Nitrospira)

Page 7: [2014.08.25] Albertsen ISME15 CAMI: Why metgenomics is broken

CENTER FOR MICROBIAL COMMUNITIES | AALBORG UNIVERSITY

The simplest metagenome in

the world

Reads Assembly

Remember to correct for read abundance

99%

1 %

50% 50%

Abun

danc

e (%

)

Accumulibacter (Class Betaproteobacteria) Nitrospira (Class Nitrospira)

Page 8: [2014.08.25] Albertsen ISME15 CAMI: Why metgenomics is broken

CENTER FOR MICROBIAL COMMUNITIES | AALBORG UNIVERSITYVisualisation: Ray Cloud Browser (S. Boisvert)

Assembly

... of the simplest metagenome in the world

Page 9: [2014.08.25] Albertsen ISME15 CAMI: Why metgenomics is broken

CENTER FOR MICROBIAL COMMUNITIES | AALBORG UNIVERSITY

Coverage (Abundance)Reads: 2x250 bp MiSeqAssembly: CLC [kmer 64]

ANI

Page 10: [2014.08.25] Albertsen ISME15 CAMI: Why metgenomics is broken

CENTER FOR MICROBIAL COMMUNITIES | AALBORG UNIVERSITY

Sequencing

Assembly

Genome Binning

Re-assembly

Manual curation

Page 11: [2014.08.25] Albertsen ISME15 CAMI: Why metgenomics is broken

CENTER FOR MICROBIAL COMMUNITIES | AALBORG UNIVERSITYL. Sauder & J. Neufield, University of Waterloo (in prep)

Page 12: [2014.08.25] Albertsen ISME15 CAMI: Why metgenomics is broken

CENTER FOR MICROBIAL COMMUNITIES | AALBORG UNIVERSITYL. Sauder & J. Neufield, University of Waterloo (in prep)

Assembly statistics

# Scaffolds 447

Total Mbp 4.46

% GC 33.7

N50 22281

Essential genes 31/29

High-quality draft genome by HMP

standards

Page 13: [2014.08.25] Albertsen ISME15 CAMI: Why metgenomics is broken

CENTER FOR MICROBIAL COMMUNITIES | AALBORG UNIVERSITYL. Sauder & J. Neufield, University of Waterloo (in prep)

Page 14: [2014.08.25] Albertsen ISME15 CAMI: Why metgenomics is broken

CENTER FOR MICROBIAL COMMUNITIES | AALBORG UNIVERSITYL. Sauder & J. Neufield, University of Waterloo (in prep)

Taxonomic classification of essential genes

Scaffold Length

Page 15: [2014.08.25] Albertsen ISME15 CAMI: Why metgenomics is broken

CENTER FOR MICROBIAL COMMUNITIES | AALBORG UNIVERSITYL. Sauder & J. Neufield, University of Waterloo (in prep)

Taxonomic classification of essential genes

Scaffold Length

All A B C# Scaffolds 447 148 135 164

Total Mbp 4.46 1.40 1.48 1.58

% GC 33.7 34.2 34.2 33.5

N50 22281 17655 27290 21230

Essential genes 31/29 26/29 2/29 3/29

High-quality draft genome by HMP

standards

AB

C

Page 16: [2014.08.25] Albertsen ISME15 CAMI: Why metgenomics is broken

CENTER FOR MICROBIAL COMMUNITIES | AALBORG UNIVERSITYL. Sauder & J. Neufield, University of Waterloo (in prep)

Abundance

Scaffold Length

Page 17: [2014.08.25] Albertsen ISME15 CAMI: Why metgenomics is broken

CENTER FOR MICROBIAL COMMUNITIES | AALBORG UNIVERSITYL. Sauder & J. Neufield, University of Waterloo (in prep)

Abundance

Scaffold Length

Page 18: [2014.08.25] Albertsen ISME15 CAMI: Why metgenomics is broken

CENTER FOR MICROBIAL COMMUNITIES | AALBORG UNIVERSITYExample dataset in the R package

Page 19: [2014.08.25] Albertsen ISME15 CAMI: Why metgenomics is broken

ScaffoldTaxonomic

classification

Tetranucleotide frequencies

Presence of key genes or pathways

% GC

Length

Abundance in different samples

Read connections to other scaffolds

CENTER FOR MICROBIAL COMMUNITIES | AALBORG UNIVERSITY

In situ proximity

Page 20: [2014.08.25] Albertsen ISME15 CAMI: Why metgenomics is broken

mmgenomeA toolbox for reproducible genome

extraction from metagenomes

goo.gl/ohCVPgMadsAlbertsen/mmgenome

CENTER FOR MICROBIAL COMMUNITIES | AALBORG UNIVERSITY

Page 21: [2014.08.25] Albertsen ISME15 CAMI: Why metgenomics is broken

CENTER FOR MICROBIAL COMMUNITIES | AALBORG UNIVERSITY

Unknown unkowns“We pray that micro-diversity didn’t compromise

the assembly and hope our key genes are present.”

Page 22: [2014.08.25] Albertsen ISME15 CAMI: Why metgenomics is broken

CENTER FOR MICROBIAL COMMUNITIES | AALBORG UNIVERSITY

goo.gl/ohCVPg

@MadsAlbertsen85

slides: goo.gl/xxxxxAcknowledgements

J. NeufieldL. Sauder

G.W. Tyson

C. Dorringer H. Daims

P. Hugenholtz

University of Vienna

Per H. NielsenSøren M. KarstRasmus H. KirkegaardSimon J. McIlroy

K. NelsonS. IshiiS. Ishii

A. Loy M. Wagner F. Moeller

T. Rattei

A. Schramm M. Lund

M. KuhlL. Behrendt

M. Sommer C. Munch