Figure 5
From: Identification of novel DNA repair proteins via primary sequence, secondary structure, and homology
Correlation of genome size to number of repair proteins and repair genes. The outer plot shows the relation between the size of a genome (horizontal axis) and the number of predicted repair-related protein transcripts in that genome (square points). Dual thresholds are used for Methods P (red), PH (green), and BLAST (blue), where one threshold produced no false positives in identification experiments (lower three curves), and another threshold produced a small percentage of false positives while admitting more true positives (upper curves). In the small inner plot, we reduce the number of detected proteins to the number of unique genes (vertical axis), and again plot versus the genome size (triangular points). The following thresholds are used for each detection technique: Method P: 0.001/0.151; Method PH: 0.001/1.015; BLAST:10-3/10-9. Data is smoothened by using bezier curves. Correlations between the size of a genome and the number of repair transcripts are listed in Table 8 for both of the Method P thresholds used. Though the tighter threshold dataset of the inner plot visually appears to be approximated by a constant for methods PH and BLAST, the slope of the approximation is positive.