HR: 0800h
AN: B11C-0619    [Abstracts]
TI: Generalized Logical Network Modeling of Interactions Among Bacteria in Aerosols Under Meteorological Factors Using High Density Phylogenetic Microarrays
AU: Song, J
EM: joemsong@cs.nmsu.edu
AF: New Mexico State University, P.O. Box 30001, MSC CS, Las Cruces, NM 88003, United States
AU: * Luce, C
EM: cluce@cs.nmsu.edu
AF: New Mexico State University, P.O. Box 30001, MSC CS, Las Cruces, NM 88003, United States
AU: DeSantis, T
EM: tdesantis@lbl.gov
AF: Lawrence Berkeley National Lab, 1 Cyclotron Road, Berkeley, CA 94720, United States
AU: Arkin, A
EM: APArkin@lbl.gov
AF: Lawrence Berkeley National Lab, 1 Cyclotron Road, Berkeley, CA 94720, United States
AU: Brodie, E
EM: ELBrodie@lbl.gov
AF: Lawrence Berkeley National Lab, 1 Cyclotron Road, Berkeley, CA 94720, United States
AU: Andersen, G
EM: GLAndersen@lbl.gov
AF: Lawrence Berkeley National Lab, 1 Cyclotron Road, Berkeley, CA 94720, United States
AB: The generalized logical network model utilizes temporal information in the 16S rRNA gene concentration time- course to examine interactions of bacteria within a microbial community under meteorological factors. The Biowatch aerosol bacterial community data set (Brodie et al., PNAS 104[1]:299-304, 2007) of 8,763 taxa intensities was generated using 237 16S rRNA oligonucleotide phylogenetic microarrays at 30 locations throughout the U.S. over time-courses of up to 20 weeks at each location. Each microarray contains about 9,000 probe sets, with an average of 24 probes per set. Seventy-two meteorological factors were measured at the time each microarray was analyzed. In a generalized logical network, a generalized truth table, associated with every node representing either a bacterial taxon or a meteorological factor, describes the represented bacterial behavior dictated by some environmental factors in addition to associations with other bacterial taxa. The optimal generalized logics at each bacterial node in the network will be searched so that they best explain the observed time-course data. Determination of an optimal logic will involve parent node selection and generalized truth-table generation. The maximum number of parents is set to a given number. If the current node shows consistent behavior during transition from one state to another given the parent nodes, then the parent nodes are kept. The actual goodness of the transition is calculated using the chi-square test. In addition to the dependency of the concentrations of bacteria on meteorological factors, with various time delays, the initial generalized logical network modeling results indicate that the concentrations of specific bacterial taxa are also associated with concentrations of other bacteria.
DE: 0545 Modeling (4255)
SC: Biogeosciences [B]
MN: 2007 Fall Meeting