PMC:2654804 / 23394-24520 JSONTXT

{"target":"https://pubannotation.org/docs/sourcedb/PMC/sourceid/2654804","sourcedb":"PMC","sourceid":"2654804","source_url":"https://www.ncbi.nlm.nih.gov/pmc/2654804","text":"aBest known solutions, based on comparsion to the solutions of the original problems. For ss_30genes2, the similarity with the source model strongly indicates that our solution is the best, but since we do not have access to the solution of the original problem this is not confirmed.\nb The majority of runs have an error below the error of the source model, but differ slightly in structure.\ncReal datasets with relatively few experiments and/or data-points making many similar models reasonable, increasing the variability in found solutions.\ndSince no source model is available, we have evaluated the original solution with our error function, and it has an error higher than our solution.\nThe error of the source model refers to the error of the model from which data was simulated. The best result is taken from several runs with various random seeds. The error, the negative log-likelihood and the number of false positives and negative (FP/FN) reactions are reported. Stability measures the frequency of runs for which the best structure is obtained. Computation time in seconds is given as the average of several runs.","divisions":[{"label":"footnote","span":{"begin":0,"end":284}},{"label":"p","span":{"begin":0,"end":284}},{"label":"footnote","span":{"begin":285,"end":392}},{"label":"p","span":{"begin":285,"end":392}},{"label":"footnote","span":{"begin":393,"end":544}},{"label":"p","span":{"begin":393,"end":544}},{"label":"footnote","span":{"begin":545,"end":692}},{"label":"p","span":{"begin":545,"end":692}}],"tracks":[]}