1. The bioinformatics behind
shotgun metagenomic
sequencing
Roche UG
2012
Rob
Edwards

2. Outline
Metagenomics
Annotation
Virus:host prediction

3. How Much has Been sequenced?
Environmental
sequencing
100
bacterial
genomes
First
bacterial
genome
1,000
bacterial
genomes
Number of known sequences
Year

4. How Much will Be Sequenced?
Everybody in
USA
Everybody in
the Indy 500
infield
One genome from
every species
100
people
Most major
microbial environments
All
cultured
Bacteria
Training people for the future
X-Prize competition, sequence 100 human genomes within 10 days or less, with an accuracy of no more than one error in every 100,000 bases sequenced, with sequences accurately covering at least 98% of the genome, and at a recurring cost of no more than $10,000 per genome

5. Metagenomics Sequencing the World

6. Shark Metagenomes

7. Shark Metagenomes

8. Metagenomics Analysis Steps
1. clean the data (prinseq)
Check the quality of the data
Remove bad reads
2. annotate the data (rtmg)
What is there?
Who is there?
3. analyze the data

9. Clean, dereplicate, check quality, analyze
Prinseq
Clean, dereplicate, check quality, analyze
PRINSEQ
PRINSEQ Results
Rob

10. Data Processed May 2012 Datasets processed: 6,223
Sequences processed: 8,698,402,976
Bases processed: ,472,118,033

11. Length Matters

12. Preprocessing Data

13. Annotating Metagenomes
Identify functions // organisms present in the sample
BLAST is very slow
Immediate processing of data
RTMG

14. A Better Way To Annotate Metagenomes
Real time metagenomics
RTMg
RTMG

15. Shark Microbe Metabolism
Microbes on sharks struggle for iron

16. Host – Virus Interactions
Can we predict the host a virus infects
just from the sequence?
Bas
Michiyo

17. How To Predict Hosts Count all 2-, 3-, 4-, 5-, 6-, 7- bp sequences
AA, AG, AC, AT, GA, GG, GC, GT …
AAA, AAG, AAC, AAT, AGA, AGG, AGC, AGT …
AAAA, AAAG, AAAC, AAAT, AAGA, AAGG, …
Count in host and virus
Use machine learning (Random Forest) to identify which hosts and which viruses match

18. Classification Accuracy
Using all known viruses and their hosts
Oligonucleotide length
Classification Error %

19. Length Matters! Using all known viruses and their hosts 200 bp reads
Using known samples as control:
~5% of reads classified
Sequencing error little effect 50 40 30 20 10
Correct predictions
Wrong predictions
Prediction percent

20. Predicted host
Actual host

21. Predicted Actual 89 % of misclassifications are near-neighbors
11% are outside the near neighbors

22. Shark virus Hosts
Virus hosts include eukaryotes, bacteria, and plants

23. Thanks
Liz
Forest
Stuart
Katie
Alan

24. Take Home Points Check your data (e.g. prinseq)
Annotate the data (e.g. RTMg)
Analyze your data

25. The Lab Ramy Jeremy Bas Dave Sajia Rob Kate Joakim Brad Steve Sheridan
Stephanie
Adam
Carny
Rima
Josh
Daniel
Michiyo
Vasken
Matt S
Matt H
Bianca
Andrés
Nick C
Nick T
Brian
Geni
Jimmy
Amanda

26. Funding PhAnToMe TUES Viral Dark Matter Brazil-US Marine
Sciences Consortium
Coral Reef
Image Analysis