The erratum and corrigendum address specific errors in the legends of figures from two articles published in *Nature Biotechnology* in 2003. 1. **Development of Potent Monoclonal Antibody Auristatin Conjugates for Cancer Therapy**: - The legend to Figure 4d on page 781 should have read "cAC10 Ag−" instead of "clgG Ag−." 2. **A Model of Molecular Interactions on Short Oligonucleotide Microarrays**: - The legend to Figure 1 on page 819 was incorrect, with parts b and c transposed. The correct legend should read: (b) Weight factors; (c) Nearest-neighbor stacking energy, which weakly correlated (r = 0.6) with those found in aqueous solution and are smaller in magnitude. - The legend to Figure 2 on page 820 was also incorrect, with figure parts referred to incorrectly. The correct legend should read: Accuracy test, comparing known concentrations of 14 'spike-in' genes with those obtained from PDNN, MAS5.0, and dChip. The microarray raw data were from the '1532 series' human data. Figures d, e, and f show standard deviations versus the averages of log-transformed expression levels for genes other than the 'spike-ins,' using PDNN, MAS5.0, and dChip, respectively. Each figure contains 12,474 genes, with the top half shown in red.The erratum and corrigendum address specific errors in the legends of figures from two articles published in *Nature Biotechnology* in 2003. 1. **Development of Potent Monoclonal Antibody Auristatin Conjugates for Cancer Therapy**: - The legend to Figure 4d on page 781 should have read "cAC10 Ag−" instead of "clgG Ag−." 2. **A Model of Molecular Interactions on Short Oligonucleotide Microarrays**: - The legend to Figure 1 on page 819 was incorrect, with parts b and c transposed. The correct legend should read: (b) Weight factors; (c) Nearest-neighbor stacking energy, which weakly correlated (r = 0.6) with those found in aqueous solution and are smaller in magnitude. - The legend to Figure 2 on page 820 was also incorrect, with figure parts referred to incorrectly. The correct legend should read: Accuracy test, comparing known concentrations of 14 'spike-in' genes with those obtained from PDNN, MAS5.0, and dChip. The microarray raw data were from the '1532 series' human data. Figures d, e, and f show standard deviations versus the averages of log-transformed expression levels for genes other than the 'spike-ins,' using PDNN, MAS5.0, and dChip, respectively. Each figure contains 12,474 genes, with the top half shown in red.