Introduction to Bioinformatics and Computational Genomics

Week 12Microbiomes and microbial communities

Most bacteria have never been cultured. Sequencing lets you study them anyway, and almost all of the analysis is arithmetic on one table.

Questions this week answers

  • Why sequence a community at all instead of growing it on a plate?
  • Clustering at 97% identity is a convention. What does it approximately buy you?
  • Two samples both contain six taxa. Are they the same kind of community?

By the end of this week you can

  • Explain the great plate count anomaly and why it forced sequencing-based methods
  • Choose between amplicon and shotgun sequencing, and justify the cost
  • Compute and interpret alpha diversity and beta diversity
  • Say what a 97% OTU clustering threshold approximates
0 of 4 done

  1. Richness counts the species; Shannon notices which ones matter

    Hold the number of taxa fixed and move the abundances around. Richness will not budge. Everything else will.

    One taxon holds 62% of the reads. Richness still says 6, and Shannon has fallen to 1.03 against a possible 1.79. This is the community a richness count flatters most.

    Shannon (H')1.03a perfectly even 6-taxon community reaches 1.79
    Richness6taxa present at all
    Evenness0.57H' over its maximum
    Most abundant taxon62%780 reads in total
    Reads per taxon6 taxa
    480
    210
    45
    30
    12
    3
    Set a taxon to zero to remove it from the community entirely.
    Things to try0 of 3

    Definitions and the Shannon formula are from the week 12 microbiome lecture.

What should survive this week

  • Most bacteria have never been cultured. In seawater, culturable cells are 0.001 to 0.1% of the total, which is what forced sequencing-based methods.
  • Amplicon sequencing is cheap and tells you who is there. Shotgun is expensive and tells you what they could do. Function predicted from 16S is predicted, not measured.
  • 97% identity approximates genus. The 16S gene is too conserved to resolve species reliably, and that is the trade you accept for universal primers.
  • Richness counts taxa; Shannon weights them by abundance. Alpha diversity is within a sample, beta diversity is between samples.