Informal education

Informal education

BSc Forensic & Analytical Chemistry (1992-1996)

• Mostly chemistry (1yr industry), final year forensics
• Final year honours project: proteins and computers
• Sinclair Basic and Frontier: Elite II

Body Fluid Analysis (1996)

• Glasgow Royal Infirmary

Making the switch

PhD (1996-1999)

• Evolution of snake venom toxins (correlated mutations)
• Neural network model of protein evolution
• Drug active site discovery algorithm (still in use)

doi:10.1006/jmbi.1998.2437; doi:10.1006/jtbi.1999.1043; PMID:11579223; PMID:12676977

First steps in academia

Postdocs (1999-2003)

• Systems biology: modelling yeast metabolism
• Directed evolution: improved lanthanide binding

doi:10.1046/j.1432-1033.2002.03055.x; doi:10.1016/j.jtbi.2004.12.005

Multidisciplinarity

The SCRI/JHI Years

Bioinformatician (2003-present)

• Research Institute
• Government funding, policy remits
• Ineligible for many usual funding sources (e.g. RCUK)
• Bioinformatician/Computational Biologist
• Not a clear postdoc/PI distinction
• BA Mathematics (Open University)

Bioinformatician (2003-present)

• Bacterial genomics
• Plant-microbe interactions
• Oomycete genomics
• Microbial systems biology
• Microbial synthetic biology
• (occasionally plants, fungi, nematodes and viruses)

2003: Erwinia

• Arrived at SCRI part-way through sequencing Erwinia carotovora subsp. atrosepticum

(http://sulab.org/2013/06/sequenced-genomes‐per‐year/)

Global pathogens

• Blackleg, stem-rot
• Quarantine pathogens

Toth et al. (2011) doi:10.1111/j.1365-3059.2011.02427.x

First enterobacterial plant pathogen genome

doi:10.1073/pnas.0402424101

Acceptable in the 00s…

• 32-author single bacterial genome paper!
• £250,000 collaboration between SCRI, University of Cambridge, Wellcome Trust Sanger Institute
• All repeats and gaps bridged and sequenced directly
• A single complete high-quality 5Mbp circular chromosome
• 3 person-years’ manual annotation

TIL: Annotation

• The utility of genomes depends on annotation
• Annotation is curation, not cataloguing
• Automated annotation from curated data (e.g. with Prokka) now the only game in town
• But you can’t propagate what doesn’t exist
• Lots of genomes, few incentives to curate well: “many parents, but no-one wants to look after the children”
• Centralised data resources (e.g. PhytoPath) only a partial solution (funding?)

TIL: Comparisons

• The real power of genomics is comparative genomics
• \(\textrm{genome} \implies \textrm{heritable}\)
• identification of functional elements, evolutionary processes and constraints
• But epigenetics, tissue differentiation, mesoscale, phenotype plasticity, …

Comparisons

• So we compared Pba to the 130 prokaryotic genomes available at the time…

doi:10.1146/annurev.phyto.44.070505.143444

Functional differences

• Pba-only: pathogenicity determinants; phage/IS elements
• Pba & environmental: pathogenicity determinants; surface proteins; regulatory proteins

Visualisation

GenomeDiagram

doi:10.1093/bioinformatics/btk021

Biopython/KGML

http://biopython.org

Art/Science

• Elaine Shemilt, exhibited Singapore, Dundee, London

A Blueprint for Bacterial Life and Art

Bacterial genomics II

SPI-7 & cfa

• 11 horizontally-acquired islands: one similar to a P. syringae phytotoxin synthesis island (payload swapped)

SPI-7 & cfa

• Island present in Pba, some Pcc, no Dickeya
• cfl and cfa7 Tn mutants showed reduced virulence (17dpi)

Slawiak & Lojkowska (2009) doi:10.1007/s10658-008-9418-7

2013: Dickeya

• Genomics commodified (several services used)
• 25 genomes, no fanfare, minor publications

• 6 species, draft genomes, automated annotation

doi:10.1128/genomeA.00978-13 doi:10.1128/genomeA.00087-12

Quarantine

• Dickeya: global pathogens

Diagnostic qPCR primers

• Genomes enable bulk design

doi:/10.1371/journal.pone.0034498
https://github.com/widdowquinn/find_differential_primers

Tangled Taxonomy

• 3/4 Dickeya spp. misclassified at NCBI (MiSI: 18%)

doi:10.1111/j.1365-3059.2012.02678.x

Nomenclature

• Old nomenclature/species: polyphasic, phenotypic
• Binomial nomenclature not designed for large amounts of genomic data, metadata curation (HGT of function…)
• Historical Dickeya renaming: collections, databases not updated

Legislation

• A political issue: legislation/policy relies on binomial nomenclature

DDH

• DNA-DNA hybridisation the ‘gold standard’ for classification

ANI

• Average nucleotide identity (ANI) ≈ in silico DDH

PYANI

• Python module and scripts

Dickeya

• Nine species-level groups (two novel)

Genomic classification

Baltrus (2016) doi:10.1016/j.tim.2016.02.004

ANIm Graphs

• Numeric matrices (e.g. ANIm scores) define graphs/networks

Graph decomposition

• Break down graphs (min. coverage, %identity)
• Identify unique cliques

• ‘genus’, ‘species’, ‘clonal’ groupings

doi:10.6084/m9.figshare.4810867.v4

Reclassify genus

• 50%+ of genome aligns, ≈85% identity

Overall classification

• Additional species reclassifications for SRE

Phyto-threats

Forest Research, Centres for Ecology and Hydrology, University of Edinburgh, University of Worcester

Phytophthora spp.

• Phytopthora spp. are devastating pathogens
• Crops routinely treated - woodlands aren’t, but trees also susceptible to imported disease
• Sudden oak death; juniper (P. austrocedri), etc.
• Phyto-threats: evolutionary genomics to inform nursery practice
• Metabarcoding of Phytophthora communities in nurseries and surroundings
• Existing identification by ITS1, but tens of diverse ITS1 sequences per genome…

SysBio and SynBio

Dickeya Metabolism

• Presence/absence of KEGG reactions
• Clues to host range and community function

Flux-balance analysis

• Whole-organism flux, prediction of KO effects
• Substrate usage, pathway optimisation (SynBio)

Pathways and host range

• Pathways, function, correspond to Dickeya spp.

• Differences in substrate usage

Food or Fuel?

• 1G: food crops; 2G: cellulosic crops; Ag waste?

• Maize stover, straw, sugarcane bagasse etc.

Dickeya for bioprocessing

• Plant Cell Wall Degrading Enzymes (PCWDEs)
• Enzymes engineered for bioethanol production, expressed in E. coli
• SynBio is a platform technology
• Engineered PCWDE libraries for SynBio?

Natural diversity

• Up to 75 CAZyme families in SRE
• Mine genomes for more variants

Generating diversity

• Gene-shuffling/directed evolution
• Exploit/recombine natural diversity
• Obtain novel structures

Protein sectors

• Correlated mutations (return of the PhD!)
• Decomposition of structure into sectors

Halabi et al. (2009) doi:10.1016/j.cell.2009.07.038

Positional epistasis

• Context-dependence of function and mutation: epistasis
• Saturating substitution of specificity sites

• McLaughlin et al. (2012) doi:10.1038/nature11500

Long way to go

Conclusions

Microbial Agrogenomics

• Diagnostics & epidemic tracking by sequencing (MinION)
• SysBio: plant-microbe interactions, phytobiome (plant and associated community)
• Integration of models and datasets is still challenging
• SynBio: engineering new response modes into crops (resistance, N-fixation)

As much as I know…

• Do work you enjoy
• Do it well
• Work with good people
• Read widely

Acknowledgements