To access the data, click or select the words “Appendix 2.” i) Variable importance in projection (VIP), ii) Hotelling’s T2, and iii) distance to Model. It showed examples of analysis results from positive and negative ionization analysis: (a) VIP of samples analyzed by positive electrospray (ESI) separated by nano-C18 column; (b) VIP of samples analyzed by negative charge ionization separated by hydrophilic interaction cyano (HILIC CN) column; (c) Hotelling’s T2 range of samples analyzed by positive ESI separated by nano-C18 column; (d) Hotelling’s T2 range of samples analyzed by negative charge ionization separated by HILIC CN column; (e) distance to Model analyzed by positive ESI separated by nano-C18 column; and (f) distance to Model analyzed by negative charge ionization separated by HILIC CN column. The green bar in (a) and (b) indicates the VIP level of each explored marker. The black color Aq_CTL_H and Aq_CTL are the aqueous humor samples from the diabetic cataract patients and the red color Aq_DM_H and Aq_DM are the aqueous humor samples from the diabetic macular edema patients in (c) and (d). The Hotelling’s T2 range test verified the homogeneity of population the samples. The T2 Crit(95%) and T2 Crit(99%) indicates 95% and 99% confident levels of Hotelling’s T2 range test respectively. The samples in (c) and (d) are within 95% confident level. It indicates the population in the two assays were homogeneous. The graphs in (e) and (f) also indicate the overall homogeneity of the samples.