MMG111 - LG NOTES

Describe various “scales” of microbes how this perspective impacts microbiome ecology

Microbial scales

1. Microbial Scales: Relative Sizes

2. Microbial Scales: Numbers

   1. Absolute vs. Relative abundance

3. Microbial Scales: Diversity

4. Microbial Scales: Unknown Diversity

5. Microbial Scales: Physiology

   1. Aerobic respiration

   2. Anaerobic respiration

   3. Fermentation

   4. Methanogens

   5. Syntrophs

   6. O₂, H₂, CO₂, CH₄

   7. C1 compounds

   8. SCFAs

   9. Sugars and polymers

6. Microbial Scales: Time

7. Microbial Scales: Genetics

8. Microbial Scales: Interactions

Explain some scientific reasoning cognitive flaws, biases, fallacies and how these may impact microbiome research study design and interpretation

• Logical fallacies, cognitive biases that can occur in microbiome research

  • 1. Association interpreted as causation

    • Causal fallacy: correlation not causation

    • Q: What alternative causal variables remain?

  • 2. Selective use of evidence

    • Confirmation bias: only look for evidence for something that we already think or suspect

    • Cherry Picking: intentionally or unintentionally choosing data that supports a hypothesis

    • P-hacking: only reported “significant” p-value correlations

    • Q: Were all outcomes and analyses reported?

  • 3. Post-hoc patterns presented as predictions

    • Post-hoc: dangers of finding your hypotheses in the data; patterns presented as predictions

    • TEXAS SHARPSHOOTER: the investigators defined the pattern and assign biological significance after examining the data

    • HARKING (Hypothesizing After Results are Known)

    • Q: Was the hypothesis specified before the analysis?

  • 4. False dichotomies and circular reasoning

    • False dichotomy: a logical error where an argument presents only two choices as the only possibilities, while hiding or ignoring other valid options

    • Circular reasoning: a logical error where an argument uses its own conclusion as proof

    • Q: Was dysbiosis independently defined and measured?

  • 5. Generalizing beyond the evidence

    • Q: Does the result replicate in another population or study?

Describe the process of hypothesis “falsification”

• The core scientific practice of trying to prove a hypothesis wrong rather than trying to prove it right

• “Process of elimination” and not just repeated verification

• Karl Popper

  • You are not proving a hypothesis, you are trying to falsify them. The one that has not been falsified is the one that’s essentially left standing.

    • Ruling out 

Define the elements and purposes of sections of a scientific paper

• Introduction: what’s known/unknown? What’s the premise and hypothesis? Overall research question.

• Methods: approaches, study design, reproducibility

• Results: data/evidence/observations. Figures and Figure Legends. Each figure should be a “sub” research question with data

• Discussion: interpretation of results

• Conclusions: what are the claims based on the evidence?

• Abstract: paper in miniature (one sentence for Intro, methods, etc.)

• Title: often the “headline” or primary claim

• Hard for humans to be objective, reason to separate the results section from the discussion/conclusion section. So maybe you won’t agree with the conclusion, but still believe the data.

Distinguish observations, interpretations, hypotheses, and causal claims.

• Observations: direct, objective facts gathered by the senses, interpretations assign meaning or value to those facts

• Hypotheses: propose testable explanations for them

• Causal Claims: assert that one variable directly forces a change in another

• Hypothesis vs Claims and Observations vs. Interpretations

  • Hypothesis is a prediction made before while a claim is a conclusion made after

  • Observation is a factual result while an interpretation contains opinion

Identify common errors and biases in microbiome research

• Logical fallacies, cognitive biases

  • 1. Association interpreted as causation

  • 2. Selective use of evidence

  • 3. Post-hoc patterns presented as predictions

  • 4. False dichotomies and circular reasoning

  • 5. Generalizing beyond the evidence

• Experimental/Study Design Problem

  • 
    

  Experimental/Study Design Problem	Example of Interpretation Issue
  No control group	False connection of the outcome to the intervention
  Participants knew the treatment	Expectations and placebo effects
  Self-reported outcomes	Measurement validity
  No dietary adherence data	Unclear whether the intended exposure occurred
  Unmeasured medication and supplement use	Potential confounding variables can impact outcome
  No power calculation	Limited statistical ability to detect a credible effect
  Microbiome not measured	Thus cannot validate “microbiome restoration”
  Intervention insufficiently described	Challenges ability to reproduce experiments

Compare observational and experimental study designs

• Observational studies

  • Cross-sectional study

    • Take a specific sample at a specific time without any follow-up

    • Single time point, looks at prevalence

  • Cohort

    • Participants are grouped by exposure and evaluated for a subsequent outcome. Compare exposed and unexposed groups

    • Several time points, comparison between groups with a trait to groups without that trait, outcome driven

  • Case-control study

    • Begin with outcome status and evaluate previous exposures

    • Looks at prior exposure, comparison between groups with a trait to groups without that trait, outcome driven

  • Randomized trial

    • Assigned to an intervention

• Experimental Studies

  • Designing a study and actually doing it

  • Clinical trial: determine effect of an intervention compared to another intervention or placebo

    • Randomized controlled trial: participants put in groups by chance, strongest design for estimating the effect of an intervention

      • Randomization helps mitigate confounding variables

    • Phase #1 SAFETY: Is treatment safe?

    • Phase #2 EFFICACY: Is treatment effective?

    • Phase #3 CONFIRMATION: How does treatment compare to standard practice?

    • Phase #4 FOLLOW UP: Does treatment work for other patients?

  • Can involve lab animals and humans

  • Clinical trials: determine effect of an intervention compared to another intervention or placebo

  • Randomized controlled trials: provide the strongest design for estimating the causal effect of an intervention, provided that allocation, adherence, outcome measurement, and attrition are adequately controlled

Identify controls, measurements, and alternative hypotheses needed to evaluate a microbiome claim

• Control: estimates what would happen without intervention

• Measurements: reported outcomes

• Alternative hypotheses: a claim in statistics that contradicts the null hypothesis

Distinguish a claim, evidence, and interpretation of research

• Claim: arguable statement or conclusion that answers a question

• Evidence: raw or processed data, measurements, or observations gathered to test that statement

• Interpretation: logic that explains what the evidence means and why it supports–or fails to support–the claim

Identify types and features of observational and experimental study designs

• Observational studies

  • Cross-sectional study

    • Take a specific sample at a specific time without any follow-up

  • Cohort

    • Participants are grouped by exposure and evaluated for a subsequent outcome. Compare exposed and unexposed groups

  • Case-control study

    • Begin with outcome status and evaluate previous exposures

  • Randomized trial

    • Assigned to an intervention

• Experimental Studies

  • Can involve lab animals and humans

  • Clinical trials: determine effect of an intervention compared to another intervention or placebo

  • Randomized controlled trials: provide the strongest design for estimating the causal effect of an intervention, provided that allocation, adherence, outcome measurement, and attrition are adequately controlled

Distinguish biological replicates, technical replicates

• Biological replicates: independent biological samples

  • Capture variation among people and support people and support population-level inference

  • Biological N = 3

  • Biological replicates look at variation based on biology

    • Correlated to statistical power, with more biological replicates increasing the statistical power of an experiment

• Technical replicates: repeated measurements of one sample 

  • Measure variability introduced by extraction, PCR, sequencing, or instrumentation

  • Improve measurement precision, but do not create independent biological observations

  • Biological N = 1

  • Technical replicates look at variation based on technique

Recognize controls, confounding effects, batch effects, and inadequate statistical power

• Control: estimates what would happen without the intervention

• Confounding effects: external factor in an experiment that affects both the independent and dependent variables; it can introduce error or bias, making it difficult to determine if the observed effects are solely due to the independent variable; controlling for them is essential to establish true causal relationships

• Batch effects: systematic difference among samples caused by when, where, or how they were processed rather than by the biological variable being studied

  • Batch effects are caused by unintended variation between sample groups based processing time, conditions, or method

    • Make sure you can identify common technical contributors

    • Comparable methods and balancing batches helps reduce batch effects

• Inadequate statistical power: can not determine what a study can detect; power depends on effect size, biological variability, significance threshold, and sample size

Decide whether a study supports association, causation, or a more limited conclusion

• To determine whether a study supports correlation or causation, consider its design, whether there are any confounding variables not tested for, and if it displays any bias or fallacies

  • Important to note, observational studies can only look at correlation and are incapable of determining causation

Explain general differences between various ways microbes make ATP

• Main ways

  • Aerobic respiration

  • Anaerobic respiration

  • Fermentation

• Catabolism: process of transforming chemical fuels such as glucose into ATP (energy)

  • 1. Substrate level phosphorylation

  • 2. Oxidative phosphorylation - chemiosmotic theory

  • Ex.

    • Glycolysis

    • Aerobic respiration

    • Anaerobic respiration

      • Ex. Lactic acid fermentation

Compare aerobic respiration and anaerobic respiration (electron donors, acceptors, products, substrates, ATP yield)

• Aerobic respiration = organic/inorganic donor AND O₂ acceptor

• Anaerobic respiration = organic/inorganic donor AND acceptor NOT O₂

• Fermentation = organic donor/acceptor

  • Lower ATP yield per cycle, but at a faster rate than cellular respiration

  • Alcoholic fermentation produces ethanol + CO₂

  • Lactic acid fermentation produced lactate

Contrast fermentation and respiration (electron donors, acceptors, products, substrates, ATP yields)

• Fermentation

  • Energy-yielding reactions that convert carbohydrates or sugars into acids and/or alcohols and/or CO₂ without oxygen (anoxic)

    • Sugars 

      • Different structures correspond to different names

      • Counting Carbons!

        • Hexose has 6 carbons

  • Fermentations are anaerobic redox processes - ATP is usually generated by substrate-level phosphorylation

• Cellular respiration breaks down glucose to release approximately 18-20 times more ATP than fermentation

• The rate of release of ATP by cellular respiration is slower than the rate of release of ATP by fermentation

• Respiration vs, Fermentation

  • Respiration = high efficiency furnace, but costs energy to run

  • Fermentation = charcoal grill slow burning, doesn’t require much energy to start

• Respiration = glycolysis, TCA (Krebs cycle), and ETC

• Fermentation uses SLP, and the modules are glycolysis and fermentation

Compare and contrast different types of fermentations with respect to substrates and products

• Sometimes you can have more than one end product from fermentation, or you might have more than one end product

  • Branched  or dismutation - several end products

  • Linear - might have one substrate, one product

• Fermentation (lactate)

  • Terminal Electron Acceptor - PYRUVATE

• ETHANOL FERMENTATION

  • Glycolysis, pyruvate fermentation to acetaldehyde and later to ethanol + CO₂ (no ETC)

  • Acetaldehyde receives electrons from NADH and is reduced to ethanol

  • Terminal Electron Acceptor - ACETALDEHYDE

  • Ex. Yeasts (Saccharomyces cerevisiae, Candida albicans)

  • Net: Glucose + 2ADP + 2Pi → 2 ethanol + 2 CO₂ + 2ATP

• Fermentations in the Human Microbiome

  • Diverse Substrates 

    • Cellulose and hemicellulose

    • Mucin

    • Sugars

    • Fatty acids

    • Proteins

  • Diverse Products

    • Fatty acids

    • Alcohols 

    • Aldehydes

• One glucose substrate and a diversity of fermentation products

  • You can get different products from the same substrate, and you can have more than one product from the same substrate in fermentation

  • Ethanol + CO₂ - two end products while Lactic acid is one end product

  • There can be multiple pathways depending on certain conditions

• Many Fermentations Produce ACIDS

• Many Fermentations produce ALCOHOLS

• Many Fermentation produced GASES

• Fermentation of Amino Acids

  • The Stickland reaction allows for the fermentation of amino acids

    • Need two amino acids, one as the electron acceptor and one as the electron donor

    • Result is that you have some different end products, depending on the kind of amino acids being fermented

    • Amino acid fermentation

  • Amino acid pairs are used as electron donors and acceptors

  • Diverse substrates 

    • Proteins

    • Peptides

    • Amino acids

  • Diverse products

    • Phenols

    • Indol

    • NH₃

    • H₂

Compare/contrast methods for identification of microbes (growth, microscopy, taxonomy and phylogeny)

• Growing Microbes

  • Essential components for growth

    • Components required to generate energy and also for biosynthesis - all about raw material, electron donor, electron acceptor, and a carbon source (ex. glucose)

    • What do cells need to do to grow? They need to be able to make energy, and they need raw materials to build. 

  • Selective Enrichment Technique

• Stains

• Evolutionary Relationships / “Geneologies”

  • Problems 

    • Few distinguishing morphological features in microbes as compared to microbes (plants, animals, etc.)

    • Microbes have been traditionally classified taxonomically (based on physiology, etc.)

    • Growth requirement: “need to grow it to know it”

• Cultivation-Independent Approaches

  • Sequencing

• Molecular Phylogeny

  • Sequence type = phylotype

  • “Candidtate” phyla = new phyla lacking cultivated representatives

  • 2 shorthands for species BUT not called species because this is based on nucleotide sequencing of rRNA

    • OTU (operational taxonomic units): groups DNA sequences into a 97% identity (rough proxy for species)

      • But if 96% may not be the same species

    • ASV (amplicon sequence variants): exact DNA sequences at single-nucleotide resolution (100% similarity)

Describe rationale for use of ssu rRNA as a “proxy” for organism phylogeny

• Conservation and variation in ssu rRNA sequence permits design of group-specific primers

• Similar “phylotypes” = similar sequences

• Choose rRNA to identify organisms 

  • Bacteria - want to compare them to each other on kind of a universal phylogenetic tree. We had to choose a gene or a process that was universal to all organisms

    • Ribosome → translation is a key component of all cellular life and are present in all forms of cellular life, bacteria, archaea, and eukarya

  • Has some conservation alternating with some regions of variation

    • Enough variation in the sequence of this gene that allows different ordinance to be classified

Define steps in ssu rRNA analysis of a single microbe

1. Isolate DNA from each organism

2. Make copies of rRNA gene by PCR

3. Sequence DNA

4. Analyze sequence

5. Generate phylogenetic tree

Describe rationale for cultivation-independent approaches in general

• Rationale for cultivation-independent approaches

  • Generally need to know something about microbes to cultivate them

  • Not all microbes readily isolated

  • Reliance on cultivation to ID or quantify microbes is not always sufficient/accurate

  • In situ biology (microbes don’t have to be “tamed”)

    • Lab conditions vs. how it survives in our gut → also not by itself because within a community of microbes

  • “Wild” microbes live with other “wild” microbes

• How to identify microbes

  • DNA of an environment and sequence

  • Any method (sequencing or metagenomic ways) to identify microbes and then count them and estimate the diversity in different samples that we can compare

  • Useful because we do not know how to grow every microbe

  • Cultivation dependent will be growing on a plate

Compare and contrast alpha diversity and beta diversity

• alpha diversity: how many taxa are in a sample?

  • Species richness = number of species in a sample

  • Species evenness = species abundances

  • Dominance = one or two species have a high abundance (dominate) the microbial community

  • Shannon Diversity Index

    • Both, but richness a bit more weighted

  • Simpson’s Diversity Index

    • Weights evenness a bit more

    • Richness - a lot of kinds

    • Evenness - equal representation of each kind

• Beta diversity: how many taxa are shared between samples?

Be able to read/interpret diversity “plots” or analyses

• PCoA

  • Relative abundance 

  • Wanting to compare the diversity of different sites 

  • Each sample has a different profile - different species, different amounts of different species

  • Each individual dot is essentially a representation of that profile

    • Each sample that has its own microbiome diversity, richness, and evenness is essentially plotted using a certain statistic

  • How to read?

    • If the two dots are close together, they have a similar microbiome profile, alpha-beta in general and also individual species and the proportions

      • Think of things being clumped together

• Relative Abundance

• Shannon Diversity Index = a formula for diversity comparisons that takes into account BOTH richness and evenness

  • Higher Shannon = Higher Diversity

• Simpson’s Diversity Index = takes into account evenness and richness but EMPHASIZES the EVENESS component thus the dominant species

  • Less sensitive to species richness

  • Higher Simpson = Higher Diversity

• Rarefaction curves show species “richness”

  • Plot number of new OTUs (or ASVs) by total number of sequences analyzed

    • What does this tell you about the diversity in A and B

      • For habitat A, need more data because we don’t know number of species

      • For habitat B it plateaus → can estimate the number of species, have enough data

• Compare and contrast alpha diversity

  • 
    

    • Left, Shannon scale

      • Dots correspond to sample environment

      • Soil seems to be higher so more diverse 

        • Top is most diverse

        • Bottom is least diverse

  • 
    

    • Different samples on x-axis

    • Soil is much more diverse

    • Remember Shannon Index - higher number means it is more diverse

• Relative vs. Absolute abundance