In vitro selection and RT-PCR
To evaluate the performance of our mRNA and cDNA display method, we performed a binding assay against anti-FLAG M2 antibody magnetic beads under neutral pH condition as described in the Materials and Methods section. For both methods, FLAG-random and 10aa-random conjugate samples were mixed to a 1:10,000 molar ratio. After the binding step and repetitive wash step, beads were recovered and directly subjected to RT-PCR. To avoid the PCR bias via over amplification, we checked the amplified product every 5 cycles (0 to 35 cycles) to determine the appropriate cycle for double-strand DNA library synthesis, (Supplementary Figure 3). Amplified DNA products from both display methods became visual at the 15th cycle, and were saturated after 20 cycles. As the selection round progressed, band intensity of the 15th cycle became stronger, implying the quantity of bead-bound conjugate products had increased over the course of selection. We also observed a non-specific band appearing around 200 bp, however this band remained weak throughout multiple rounds, thus we proceeded with the RT-PCR reaction. PCR was carried out for 20 cycles for the 1st and 2nd round libraries and 16 cycles for the 3rd round for both display methods.