So Friday was all about things we should care about…Dr. Rob Knight enlightens on the awesomeness of microbial ecology and microbiome studies and Evan (which will be posted later) regales us with tales of human genomics and structural variation in an effort to convince us the Human is still a pretty nifty animal to study!
But first…Rob
Dr. Rob Knight
University of Colorado, Boulder
Topic: Microbial Ecology and Metagenomics
“It is important to remember our world is a microbial world.”
Why you should care about microbial community ecology in the context of bioinformatics?
- Most molecular diversity can be found in microbes
- 99+% cannot be cultured, we only know then from sequences
- rRNA is an excellent/uniquely good genetic marker
- rRNA is found in all species
- Fast/slow evolving regions
- We have large preexisting databases which are good for designing universal PCR primers (from conserved regions) The Greengenes database is a really good reference for 16S rRNA.
- Microbes are ubiquitous and provide key ecosystem services: air quality, water quality, natural ecosystem balance.
So the obsession of the microbiome has dated as far back (or further) to the 1600’s. Van Leeuwenhoek made the observations that…
“…animalcules were in such enormous numbers, that all the water…seemed to be alive” –Van Leeuwenhoek (1683)
Continuing on with why we should continue to care about microbes:
- They determine whether tylenol is toxic to your liver.
- They tell you who to have sex with if you’re a fruit fly
- They steal genes from your food to help you digest it
“If you haven’t tried the sushi in St. Louis, you really shouldn’t”
“You are not a beautiful unique snowflake but your microbes might be.”
So now we can ask some questions:
- How human are we?
- How can we develop technologies that manipulate our microbiota to improve health
- How are microbes distributed across our bodies?
- How do our microbes change over time?
- Classical studies have been done and continue to be done on mice, how do we translate between humans and mice?
From a bioinformatics perspective we have a wealth of applications we can look at:
- Homology
- Sequence alignment
- Phylogeny
- Assembly
- Functional assignments
- Distance metrics
- Dimensionality reduction
- Classification
- Networks
- Data visualization
And with sequencing getting cheaper and cheaper by the day, the issue is needing to interpret vast amounts of sequence and tree data.
One of the studies highlighted by Rob was microbial biogeography on the keyboard where few microbes survive. They wanted to answer a variety of questions including if there was a Wallace line between the keys G & H. They found that each individual has a unique skin community. They used QIIME to integrate analysis of 100’s of samples!
So the QIIME workflow in short…
“Why is it called 454?…Because that’s the temperature you burn bills at when you start sequencing with it.”
They looked at specific sites of the human body, they’ve looked at infants born via vaginal versus C-section births, showing the initial microbial colonization of infants and how that microbiome changes over time as we become adults and our microbial communities stabilize.
“Do differences in microbiome matter?Ask a Fat Mouse…”
So what were they finding essentially?
- There were a lot more firmicutes in obese/obese mice
- Genetic mutation can be passed by doing a gut community transfer.
- Gut microbes associated with obesity: the more weight the mice lost the greater the change in microbial community.
Future directions look toward personalized medicine in developing nations and these pilot studies in humanized mice should allow us to look at issues of obesity and malnutrition in the developed and underdeveloped world.
