Absolute or relative transcript amounts measured through high-throughput technologies (e.g., microarrays)are now commonly used in bioinformatics analysis, such as gene clustering and DNA binding motif finding. However...Absolute or relative transcript amounts measured through high-throughput technologies (e.g., microarrays)are now commonly used in bioinformatics analysis, such as gene clustering and DNA binding motif finding. However,transcription rates that represent mRNA synthesis may be more relevant in these analyses. Because transcription rates are not equivalent to transcript amounts unless the mRNA degradation rates as well as other factors that affect transcript amount are identical across different genes, the use of transcription rates in bioinformatics analysis may lead to a better description of the relationships among genes and better identification of genomic signals. In this article, we propose to use experimentally measured mRNA decay rates and mRNA transcript amounts to jointly infer transcription rates, and then use the inferred transcription rates in downstream analyses. For gene expression similarity analysis, we find that there tends to be higher correlations among co-regulated genes when transcription-rate-based correlations are used compared to those based on transcript amounts. In the context of identifying DNA binding motifs, using inferred transcription rates leads to more significant findings than those based on transcript amounts. These analyses suggest that the incorporation of mRNA decay rates and the use of the inferred transcription rates can facilitate the study of gene regulations and the reconstruction of transcriptional regulatory networks.展开更多
文摘Absolute or relative transcript amounts measured through high-throughput technologies (e.g., microarrays)are now commonly used in bioinformatics analysis, such as gene clustering and DNA binding motif finding. However,transcription rates that represent mRNA synthesis may be more relevant in these analyses. Because transcription rates are not equivalent to transcript amounts unless the mRNA degradation rates as well as other factors that affect transcript amount are identical across different genes, the use of transcription rates in bioinformatics analysis may lead to a better description of the relationships among genes and better identification of genomic signals. In this article, we propose to use experimentally measured mRNA decay rates and mRNA transcript amounts to jointly infer transcription rates, and then use the inferred transcription rates in downstream analyses. For gene expression similarity analysis, we find that there tends to be higher correlations among co-regulated genes when transcription-rate-based correlations are used compared to those based on transcript amounts. In the context of identifying DNA binding motifs, using inferred transcription rates leads to more significant findings than those based on transcript amounts. These analyses suggest that the incorporation of mRNA decay rates and the use of the inferred transcription rates can facilitate the study of gene regulations and the reconstruction of transcriptional regulatory networks.
