Inferring gene expression from ribosomal promoter sequences, a crowdsourcing approach

摘要: The Gene Promoter Expression Prediction challenge consisted of predicting gene expression from promoter sequences in a previously unknown experimentally generated data set. was presented to the community framework sixth Dialogue for Reverse Engineering Assessments and Methods (DREAM6), effort evaluate status systems biology modeling methodologies. Nucleotide-specific activity obtained by measuring fluorescence fused upstream yellow protein inserted same genomic site yeast Saccharomyces cerevisiae. Twenty-one teams submitted results levels 53 different promoters ribosomal genes. Analysis participant predictions shows that accurate values low-expressed mutated were difficult obtain, although latter case, only when mutation induced large change compared wild-type sequence. As previous DREAM challenges, we found aggregation provided robust results, but did not fare better than three best algorithms. Finally, this study provides benchmark assessment methods specific set their sequence, it also top performing algorithm, which used machine-learning approaches, can be improved addition biological features such as transcription factor binding sites.