PhyloDetect identifies the organisms contained in a sample that was hybridized on a diagnostic microarray. PhyloDetect assumes that the microarray holds probes that can actually discriminate either individual organisms or groups of organisms. PhyloDetect does not make any assumption on the target organisms. These may be bacteria, viruses or others.
The PhyloDetect web application is designed for the Firefox browser. We will make it compatible with Internet Explorer in the future.
If you have questions please send an e-mail to Hubert.Rehrauer@fgcz.ethz.ch
For each microarray type a probe-organism match matrix (probe match example) is required. The file must be in tabular format and hold in the first column the probe identifiers and in the first row the organism identifiers. The cells of the table should be "1" if the probe matches the organism and therefore should give a hybridization signal, and "0" if the probe does not match.
PhyloDetect also allows to provide annotation data for the target organisms which is then used for display on the web pages. Annotation data has to be in the same form as this annotation example.
For each hybridization the outcome has to be loaded as a two column text file (hybridization outcome example). Again the first column must hold the probe identifiers. The second column must specify for each probe if it was classified as present ("1") or absent ("0") in the hyb.
PhyloDetect organizes the organisms in an organism tree. If there are organisms that nominally cannot be distinguished by the microarray are assigned to the same tree node. PhyloDetect computes for each node of the tree the likelihood of the hybridization data under the assumption that the node was present. Only those nodes with a likelihood above a threshold (e.g. p=0.05) are considered as potentially present.