- a contribution by Kim Halpin (halpink@live.com) -

The genomic revolution has been driven by the exponential increase in DNA sequencing capabilities together with increased affordability. The human genome project was born in the molecular revolution and the entire human genome was sequenced. It took 13 years, thousands of sequencers, and $2 billion dollars, and it generated 21 Giga base pairs of sequence. Take one step forward to the genomic revolution and another human genome is sequenced, but this time it only takes 7 days, on one sequencer, costing $3000 and generating 5-10 times more data.

The key technology underpinning the genomic revolution is next generation sequencing (NGS). Traditional sequencing is based on sequencing amplified, reasonably large sections of a genome. NGS uses instruments such as Life Technologies SOLiD^TM which sequences genomic material in millions of very small pieces, enabling a high throughput and deep sequencing. “Deep” refers to coverage over a certain segment of the genome – 60X coverage means for any nucleotide, it will be read 60 times. Sample preparation breaks the DNA or RNA into short segments that are attached to 500 million to 1 billion beads. Sequence from each bead is reported in a data file, and files can be over 100 GB.

On the market for approximately 5 years, NGS (which is sometimes referred to as second generation sequencing) already has a successor and this has a single-molecule approach. One single molecule approach is presented by Ion Torrent who have developed the first product to use semiconductor sequencing technology. In nature, when a nucleotide is incorporated into a strand of DNA by a polymerase, a hydrogen ion is released as a byproduct. The charge from that ion will change the pH of the solution, which can be detected by an ion sensor. The Personal Genome Machine (PGM™) — essentially the world's smallest solid-state pH meter — will call the base, going directly from chemical information to digital information¹.

The PGM™ offers semiconductor scalability and it currently costs less than $50,000 USD. It has a footprint that is no bigger than a desktop printer. In terms of applications, it is perfect for viral and bacterial genomes. With this platform we will see a rapid acceleration in the number of viral and bacterial sequences publicly available. It was preliminary data from DNA sequencing performed on the PGM™ which strongly suggested that the bacterium at the root of the deadly food borne outbreak in Germany was a new hybrid type of pathogenic E. coli strains².

Following the initial sequencing of the Escherichia coli O104 outbreak, nine isolates have now been sequenced by four different teams on four different sequencing platforms. (including the PGM^TM, Roche's 454 GS Junior,  Illumina HiSeq and the Illumina MiSeq)³. This crowd sourcing project is being used to annotate the genomes⁴. Crowd sourcing is when an open call goes out to a large undefined group of people, and data is sourced from this large crowd, with participation being voluntary. The different groups have been making their data publicly available and researchers have started this project to analyse and annotate the different assemblies. Researchers will use the data to generate a meta-assembly. Having data from multiple platforms could help in producing the most accurate assembly because it should compensate for any errors that might arise from a single platform³.

Because different isolates have been sequenced, the different genomes can be compared to see if there are genuine differences and to see if there are mutations that have occurred during the outbreak³. This may represent the way forward particularly for projects where a prompt answer is required. It will also ensure that the molecular epidemiology is not compromised by sequencing and data errors. 

Crowd sourcing: Is it the way for the future?

References

1. http://www.iontorrent.com/about/overview/
2. http://www.lifetechnologies.com/news-gallery/press-releases/2011/dna-sequencing-data-reveals-new-hybrid-e.html
3. Heger M. (2011) E. Coli Sequencing Prompts Crowdsourcing Project to Annotate Genomes, Enabling Platform Comparisons. In Sequence available at: http://www.genomeweb.com/sequencing/e-coli-sequencing-prompts-crowdsourcing-project-annotate-genomes-enabling-platfo
4. Location of crowd sourcing project: https://github.com/ehec-outbreak-crowdsourced/BGI-data-analysis/wiki

The proportional similarity index (PSI) or Czekanowski index is an objective and simple estimate of the area of intersection between two frequency distributions (Rosef, Kapperud et al. 1985). The PSI estimates the similarity between the frequency distributions of i.e. bacterial sub types between different reservoirs. It is calculated by:

where pi and qi represent the proportion of strains belonging to type i out of all strains typed from species P and Q (Feinsinger, Spears et al. 1981; Rosef, Kapperud et al. 1985). Click on the below animation for a visualisation of the approach. The values for PS range from 1 for the highest possible similarity to 0 for distribution with no common types. Bootstrap confidence intervals for this measure can be estimated based on the approach applied by Garrett et al (Garrett, Devane et al. 2007). This technique has also recently applied to support source attribution studies of human campylobacteriosis (Mullner et al. 2009; Mullner et al. 2010).

The occurrence of non-typable strains in a dataset requires special attention when applying this method (Rosef, Kapperud et al. 1985; Garrett, Devane et al. 2007). The assumption made by this method is that epidemiological affinity between species is proportional to the similarity between the type distributions of the species being compared. This may be incorrect since many animal isolates may not be pathogenic, even if they are identical as determined by the typing method used.  If a source also contains a high proportion of non-pathogenic strains its importance as contributor to human cases may be masked (Garrett, Devane et al. 2007). In addition some of the human cases may have originated from non included sources, i.e. by travel (Rosef, Kapperud et al. 1985).

References:

Feinsinger, P., Spears, E. E., & Poole, R. W. (1981). A Simple Measure of Niche Breadth. Ecology, 62(1), 27-32.

Garrett, N., Devane, M. L., Hudson, J. A., Nicol, C., Ball, A., Klena, J. D., et al. (2007). Statistical comparison of Campylobacter jejuni subtypes from human cases and environmental sources. Journal of Applied Microbiology, 103(6), 2113-2121.

Mullner, P., Spencer, S. E. F., Wilson, D., Jones, G., Noble, A. D., Midwinter, A. C., et al. (2009). Assigning the source of human campylobacteriosis in New Zealand: A comparative genetic and epidemiological approach. Infection, Genetics and Evolution, 9, 1311-1319

Müllner, P., Collins-Emerson, J., Midwinter, A., Carter, P., Spencer S., Van derLogt, P., et al. (2010). Molecular epidemiology of Campylobacter jejuni in a geographically isolated country with a uniquely structured poultry industry. Applied and Environmental Microbiology, 76(7), 2145-2154.

Rosef, O., Kapperud, G., Lauwers, S., & Gondrosen, B. (1985). Serotyping of Campylobacter-jejuni, Campylobacter coli and Campylobacter-lardis from domestic and wild animals. Applied and Environmental Microbiology, 49(6), 1507-1510.

It is important to clearly define the terms isolate, strain and clone to avoid confusion. Ideally in all molecular epidemiological studies these terms should be applied using the same definitions. For example "strain" and "isolates" are often used synonymously, and this may results in problems, in particular when the definition of the terms is not standardised.

The below definitions are approved by the American Society for Microbiology:

Isolate: A population of microbial cells in pure culture derived from a single colony on an isolation plate and identified to the species level.

Strain: An isolate or group of isolates exhibiting phenotypic and/or genotypic traits belonging to the same lineage, distinct from those of other isolates of the same species.

Clone: An isolate or group of isolates descending from a common precursor strain by nonsexual reproduction exhibiting phenotypic and/or genotypic traits characterised by a strain-typing method to belong to the same group.

The terms isolate, strain and clone form a hierarchy, as illustrated in the figure below.

 [This figure was taken from Zadoks et al. 2006]

Sources and Resources:

Zadoks, R. N., and Y. H. Schukken. 2006. Use of molecular epidemiology in veterinary practice. Veterinary Clinics of North America-Food Animal Practice 22:229-261.

Riley, L. W. 2004. Molecular Epidemiology of Infectious Diseases - Principles and Practices. ASM Press, Washington, DC.

Which typing method? This really is the million dollar question in molecular epidemiological research. The choice of typing approach should firstly be driven by three aspects:

1. The ecologic and evolutionary scale of the research question (e.g. outbreak investigation or long-term global spread)
2. How rapidly does the pathogen evolve (e.g. rapidly evolving RNA virus)
3. Assumption of the (genetic) method

Secondly performance criteria of available typing approaches need to be assessed. These include:

1. Typeability
2. Reproducibility
3. Stability
4. Discriminatory power

The importance of the choice of typing method is illustrated by the below graph. Assume that for example the typing approach using Marker A has a higher discriminatory power than Marker B. The discriminatory power is the average probability that the typing system will assign a different type to two unrelated strains randomly sampled in the microbial population. Depending on your choice of approach your findings will diverge, either showing that Strain 1-3 are identical or different.

[This figure was taken from Tibayrenc 1998]

This illustrates how crucial the choice of an appropriate techniques as well as the interpretation of the experimental data within a sound theoretical framework are.

It is important to underline in this context that an ideal typing system for universal use does not exist. We will explore each of the four performance criteria in detail in future articles.

Sources and Resources:

Zadoks, R. N., and Y. H. Schukken. 2006. Use of molecular epidemiology in veterinary practice. Veterinary Clinics of North America-Food Animal Practice 22:229-261.

Riley, L. W. 2004. Molecular Epidemiology of Infectious Diseases - Principles and Practices. ASM Press, Washington, DC.

Tibayrenc M. Beyond strain typing and molecular epidemiology: Integrated genetic epidemiology of infectious diseases. Parasitology Today 1998; 14: 323-329.

Struelens MJ, Members of the European Study Group on Epidemiological Markers (ESGEM) of the European Society for Clinical Microbiology and Infectious Diseases (ESCMID). Consensus guidelines for appropriate use and evaluation of microbial epidemiologic typing systems. Clinical Microbiology and Infection 1996; 2: 2-11.