�ݺ�ߣ

The indoor air microbiome in
classrooms
DeninaHospodsky, Naomichi Yamamoto, Kyle
Bibby, Dana Miller, SisiraGorthala, William Nazaroff,
Jordan Peccia

Yale University, UC Berkeley

Background

 How does the human occupant shape the indoor
air microbiome?
 Classrooms
 Defined times of occupancy and vacancy
 Constant numbers of occupants
 Public places
 Children, a susceptible population, spend
considerable amount of time in classrooms

Goals: Toidentifysources and calculate emission
rates of indoor (bio)aerosols
 Objectives:
 To combine biomolecular techniques and particle modeling to
assess the influence of the human occupant on shaping the
indoor air microbiome
 To describe and model these two processes for biological and
total particles
 To identify sources of bacteria and fungi in indoor air using
phylogentic information
 Understand how building design and operation influences the
indoor air microbiome shedding

resuspension

resuspension

Eleven sampling sites- classrooms, 3-4 days of continuous occupied
and unoccupied sampling each

Arhus
Berlin
Lyngby

New Haven
Lanzhou 1
Salinas
Lanzhou 2

Study design

 Floor dust as well as indoor and outdoor air collection
during times of occupancy and vacancy using
multistage Andersen cascade impactor in eleven
classrooms
 Quantitative data from qPCR for all samples
 Bacterial and fungal emission rate calculations
 Fungal ITS region from (almost) all air and dust samples
 Bacterial 16S region for floor dust and occupied air
samples from eleven sampling sites

OccIn Occ out UO in UO out Floor Floor Floor Floor
Location (Air) (Air) (Air) (Air) 1 2 3 4

Salinas
New Haven 1
New Haven 2
New Haven 3
Arhus
Berlin
Copenhagen
Lanzhou 1 summer
Lanzhou 2 winter
Lanzhou 1 summer
Lanzhou 2 winter

Bacteria 16S
Fungi ITS only Bacteria 16S only No sample
and Fungi ITS

Study design “metadata”

 Air temperature (minute)
 Humidity of air (minute)
 Particle mass and particle number concentration (all time averaged
and per minute)
 Bacterial and fungal number concentrations in air and on dust (all
time averaged)
 # of occupants and their activity rate (minute and time averaged
per day)
 CO2 concentration and air exchange rate (minute and time averaged
per day)
 Particle, bacterial, and fungal emission rate (all time averaged)

Time scale
1min 1hour 1day 2days 3days 4days

Data and metadata sharing

 Previous sequencing data is available on MG
RAST (and via MG RAST on MoBeDac)
 metadata documentedin SI or text of publications.
 New sequencing data will be submitted to
Qiime DB and MG RAST including metadata
 No human subjects restrictions

Sequence data analysis

Study needs:
 Easy access to
sequence data
from other
groups
 Toolsets that
allow graphical
and analytical
comparison of
samples from
our and other
groups

Costello et al, Science 2009
Lauber et al, AEM 2009

What do you hope to get out of the QIIME/VAMPS workshop?

1. Data deposition
 Clear guidelines about where to deposit sequencing data when
publication date is nearing (MoBeDac? QIIME DB? MG RAST?)
 Convenient data deposition with and without “metadata”
templates
2. Share sequence data with others
 Easy access to published sequencing data of other groups,
especially within Sloan
 Access to raw data and processed data
 Clear communication on how raw sequence data was processed
3. Effective data pooling when sequence regions do not
overlap
 Alignment tools for 16S and ITS region

 Enhanced tools for ITS sequences

�ݺ�ߣ

Peccia qiime vamps_workshop

More Related Content

Peccia qiime vamps_workshop

Editor's Notes