INTRODUCTION The fruit industry employs millions of hectares for production; of these, a significant percentage uses only female or hermaphrodite plants. The selection is carried out by morphological analysis until blooming, which represents an inefficient use of resources. Typically, well-equipped laboratories and trained personnel are required to perform DNA analysis and give valid results1,2. Devices capable of generating this information within a short timespan could contribute to a fast and effective selection of productive plants. Nonetheless, it is necessary to have high sensitivity to differentiate between samples. The objective of this work is the development of an electrochemical and an optical detection system, with high sensitivity and repeatability. With the aim to incorporate the best option in a fully integrated device that ensures an easy-to-use product, safeguarding the correct use by non-trained operators and reducing the time to obtain a result. MATERIALS AND METHODS To evaluate the detection systems a random sequence of DNA (5’-TAGTCGTCAATCCTCCGCTT) was used (50 – 540 ng/µL). The electrochemical detection design is composed by three electrodes (WE-Au, AE-Au, RE-Ag|AgCl) and 0.5 M Ru(NH3)6Cl3 as a DNA intercalating agent3 in 40 mM TAE buffer. Three electrochemical techniques (SWV, DPV and CV) were compared (Bio Logic potentiostat). The optical detection system is composed of a phototransistor (Max. Sens. @ 550 nm), a filter (520 nm) and a LED (Max. WL peak @ 455 nm) to detect the fluorescent DNA intercalator SYBRgreen. Different concentrations of fluorescein (0.005 – 10 µM) were used to evaluate the system capabilities. The system was evaluated with the DNA sequence and SYBRgreen, then compared with a fluorescence microscope (Nikon Eclipse Ti-U). Finally, both systems were tested with plant DNA samples. Genomic material was extracted (CTAB protocol4) form 6 plants (3 female/3 hermaphrodite), amplified by PCR (Thermal cycler miniPCR Amplyus) and analyzed by electrophoresis. RESULTS With the electrochemical detection system, we observed a proportional decrease in the Ru(NH3)6Cl3 signal as the DNA concentration increases, as expected. The data analysis demonstrated that SWV was the most sensitive technique with a R2=0.980, LOD 30.968 ng/µL and LOQ 103.226 ng/µL. Results showed that female plants present an average current of 1.774 μA while hermaphrodite plants had 1.694 μA, allowing us to differentiate hermaphrodite plants accurately with a 90 % confidence interval. On the other hand, in the optical detection system with a R2=0.9568, LOD 2.12 ng/µL and LOQ 7.06 ng/µL, we observed an increase in the signal as de DNA concentration increases. When the optical analysis was compared with the fluorescence microscope results, we observed a similar trend in the signals with a R2=0.995, that validate the functionality of our system. However, the difference between plant samples was detected with only the 70 % confidence interval, female samples showed an average signal of 381 mV while hermaphrodite plants had 352.56 mV. CONCLUSIONS Plant DNA detection was evaluated through two different systems. The optical detection system showed more sensitivity with synthetic samples. However, in the evaluation of plant samples, the electrochemical system is less susceptible to interferences. Therefore, we can differentiate between female and hermaphrodite plants with a 90 % confidence interval. In this way, it represents the best option to evaluate a lab-on-a-chip device. REFERENCES Barrantes-Santamaría W, Loría-Quirós C, Gómez-Alpízar L. Evaluation of two-sex determining systems in papaya plants (Carica papaya) Pococí hybrid. Agron Mesoamerican. 2019;30(2):437-446. doi:10.15517/am.v30i1.34916Cornelis S. FORENSIC LAB-ON-A-CHIP DNA ANALYSIS. 2019.Li LY, Jia HN, Yu HJ, et al. Synthesis, characterization, and DNA-binding studies of ruthenium complexes [Ru(tpy)(ptn)] 2 + and Ru(dmtpy)(ptn)] 2 +. J Inorg Biochem. 2012;113:31-39. doi:10.1016/j.jinorgbio.2012.03.008Biosciences. CTAB extraction solution for genomic DNA extraction. Biosciences. 2016:3-7. Figure 1