Computational protocol: Proteome Changes Driven by Phosphorus Deficiency and Recovery in the Brown Tide-Forming Alga Aureococcus anophagefferens

Similar protocols

Protocol publication

[…] The mass spectra collected in this study were searched using SEQUEST (Bioworks version 3.3, Thermo Inc., San Jose CA). An amino acid database for A. anophagefferens was constructed by combining all “project data” from the A. anophagefferens genome sequencing (11520 sequences from NCBI: and adding plastid proteins (105 sequences from NCBI:, along with common contaminants as well as a reversed ‘decoy’ version of these databases for false discovery rate analysis (data downloaded on March 8th, 2011). Searches were conducted with a static modification for cysteine of +57 for alkylation by iodoacetamide and allowing for variable modifications expected if methionine was oxidized (+16), if cysteine or methionine were present as seleno-residues (+47) or if selenocysteine was modified to dehydroalanine (−91) . Database search results were further processed using the PeptideProphet statistical model within Scaffold 3.0 (Proteome Software Inc., Portland OR). At least two peptides had to map to a protein sequence to be included in the data. Relative protein abundance was determined using spectral counting in Scaffold 3.0. Spectral counts are normalized across samples in each experiment, including technical replicates, to allow comparison of relative protein abundance and result in a quantitative value abundance score, as previously described . Proteins discussed as ‘differentially abundant’ were determined by the Fisher exact test as previously described with p-values<0.05. A complete list of p-values for all proteins can be found in . False discovery identification rate was estimated using a reversed decoy database as previously described .The proteins that met the criteria for being differentially abundant were compared by a hierarchical cluster analysis using Cluster 3.0 . Average abundance scores for each sample were log transformed, centered about the mean and normalized by multiplying all values by a scale factor S so that the sum of the squares of the values for each protein is 1.0. The treatments were not centered or normalized. The data were then clustered by both protein and treatment using a centered correlation as metric and complete linkage as clustering method. The data were displayed using Java Tree View . […]

Pipeline specifications

Software tools Comet, PeptideProphet
Application MS-based untargeted proteomics
Organisms Aureococcus anophagefferens
Diseases Glucosephosphate Dehydrogenase Deficiency, Phosphorus Metabolism Disorders
Chemicals Phosphates, Phosphorus