Fig. 3.

Group quality and consistency comparison. The conventional XCMS approach without Warpgroup was compared to the XCMS approach with Warpgroup. Quality of generated groups was assessed. Groups were manually inspected and rated on a scale of 0–4. Zero scores corresponded to noise groups with no discernable correct integration. The remaining scores ranged from 1 (integration regions incorporating different peaks across samples) to 4 (identical integration regions across all samples). Warpgroup (right) showed a major improvement in group quality as compared to the conventional workflow (left). Warpgroup also showed an expected increase in noise groups