Transcriptional clusters identified by expression profiling over nine conditions. The data from the 18 different arrays were normalized such that the mean average difference for all genes was 200 (approximately two copies per cell). For clustering, the signals for each gene were normalized so that the median for all conditions was one. Representative clusters are shown in a–d, including clusters in which genes are induced after treatment with methyl methane sulfonate (MMS) and ultraviolet light (UV), induced after treatment with hydroxyurea (VIII), expressed on growth in glycerol-containing media (XVI), and repressed after treatment with MMS or UV (XVIII). For highly expressed genes, the fold change is likely to be underestimated because of the nonlinear response of the fluorescence signal at high concentrations. All data can be downloaded from http://pub.gnf.org/∼ewinzeler/identification_of_new_gene.htm.

Transcriptional profile of the nonannotated open reading frame (NORF) NPR002C and the flanking neighboring genes YPR010C and YPR011C. (a) Array hybridization images. Each open reading frame (ORF) and NORF is represented on the S98 array by 16 oligonucleotide pairs. One member of each pair corresponds to a perfectly matched sequence from the ORF (PM); the other pair member contains a single-base mismatch in a central position (MM). The difference in intensity between the perfectly matched and the mismatched sequences (PM-MM) is used to calculate an “average difference intensity” for each ORF in each experiment. Array probe hybridization images for NORF NPR002C and ORF YPR011C from control cells in logarithmic phase growth, cells treated with HU, UV, MMS, and cells grown in glycerol containing media-treated cells are shown along with the average difference (Avg Diff) intensity values. (b) The average difference intensity of each gene graphed across all the conditions tested in this study. (c) Chromosomal view of NPR002C, YPR011C, and YPR010C with the distance in nucleotides between the NORF and ORF printed above the gap regions. The correlation of expression profiles between NPR002C and the upstream gene YPR011C and the downstream gene YPR010C is 0.13 and −0.32, respectively.

Northern blot analysis of NPR002C and YPR011C. (a) Expression of YPR011C across various conditions. RNA was extracted and total yeast RNA was separated by electrophoresis in an agarose gel, blotted, and hybridized with a polymerase chain reaction (PCR) amplicon of YPR011C. (b) The same blot was then stripped and hybridized with a PCR amplicon of NPR002C.

Homologs of NORF NNL005C are found in other species. CLUSTAWL alignment of homologous protein sequences from the mouse RIKEN cDNA 0610041E09 gene, Drosophila CG14199 gene, and the yeast NORF NNL005C. The mouse sequence scores (P < 8.3 × 10⁻²²) and the Drosophila sequence scores (P < 2.0 × 10⁻²⁰).

Mass spectra for a peptide from the NORF NIL001W. A multidimensional protein identification technology (MudPIT) analysis of the soluble proteome of BJ5460 was performed and the results analyzed via SEQUEST (Eng et al. 1994) using a concatenated database containing ORFs and NORFs. In the MudPIT analyses, a collision-induced dissociation tandem mass spectrum for (M + 2H) ²+ ion of the peptide DILDVLNLLK at m/z 578.5 from the NORF NIL001W was detected and identified. An eight-ion b and seven-ion y series are shown in red and blue, respectively, and the corresponding amino acid difference between each ion is shown. The SEQUEST result for the tandem mass spectrum shown had an Xcorr of 3.1276 and a ΔCn of 0.2292, indicating complete confidence in the SEQUEST result.