An analysis of simple computational strategies to facilitate the design of functional molecular information processors

Abstract

Background: Biological macromolecules (DNA, RNA and proteins) are capable of processing physical or chemical inputs to generate outputs that parallel conventional Boolean logical operators. However, the design of functional modules that will enable these macromolecules to operate as synthetic molecular computing devices is challenging.

Results: Using three simple heuristics, we designed RNA sensors that can mimic the function of a seven-segment display (SSD). Ten independent and orthogonal sensors representing the numerals 0 to 9 are designed and constructed. Each sensor has its own unique oligonucleotide binding site region that is activated uniquely by a specific input. Each operator was subjected to a stringent in silico filtering. Random sensors were selected and functionally validated via ribozyme self cleavage assays that were visualized via electrophoresis.

Conclusions: By utilising simple permutation and randomisation in the sequence design phase, we have developed functional RNA sensors thus demonstrating that even the simplest of computational methods can greatly aid the design phase for constructing functional molecular devices.

Keywords: Computational RNA; Molecular computing; Molecular logic circuit; Molecular programming; RNA computing.

Fig. 1

Conceptual representation of molecular SSD. a presents the segmentation identification of SSD comprising of 10 individual pins, where each pin can be switched (ON or OFF). Seven of the pins will correlate to the seven LED segments [Refer Additional file for the truth table of the molecular SSD]. b illustrates the distribution of logic gates per wells. Each circle represents a gate and gates are distributed accordingly into the wells based on the digits it corresponds to. Wells are labelled as 1 through 15 from left to right and from top to bottom such that the upper left well is labelled as well 1 and the lower right is labelled as well 15. c is a conceptual representation of a molecular seven-segment character display. Gates in specific wells will be activated upon binding with its input sequence. For instance, well 1, 2, 3, 6, 9, 12, and well 15 will be activated by the input sequence represent the numerical digit seven (i.e., activation of gate 7)

Fig. 2

The dependency diagram for YES gate. The colours represent the inter-binding of each base position. These coloured bases are interdependent where changing one base should change its complementary base in order to retain the secondary structure whereas the white bases indicate that the base is not complementary or binds to any base position

Fig. 3

Schematic representation of the three algorithms for logic gate design. a First strategy: the steps to permute nine bases within the OBS YES-1 gate. b Second strategy: the steps to perform random substitutions of seven bases within the OBS YES-1 gate and c Third strategy: substitutions of the complete OBS YES-1 sequence

Fig. 4

The passing rate of candidates for each criterion. Six criteria (not more than three consecutive nucleotides, present in the inactive state, 30–70 % OBS binding, diversity value not more than 9, having energy gap within -6 to -10 kcal/mol and have at least 58 % OBS GC pairing) were used to select the candidate sequences

Fig. 5

Profile of ribozyme assays visualized in 10 % denaturing PAGE at 59 V. Gate activity without the complementary input oligonucleotides (-) and with the presence of complementary input oligonucleotides (+) is showed. The smears beneath each band show immature transcripts of DNA. This does not influence the self-cleavage reaction. Figure 5 shows the self-cleavage activity for candidates generated by (a) first strategy (b) second strategy and (c) third strategy

Fig. 6

Mismatch profile for candidate gates. This figure presents the activity of the gates for the first (a) and second strategy (b) with the insertion of ten input DNA oligonucleotides that were visualized in 10 % denaturing PAGE at 59 V Input 0 through input 9 were inserted into each well respectively. Each well depicting double bands indicates the occurrence of cleavage reaction. The highlighted cells represent the gate with its complementary input sequences. Character ‘X’ indicates mismatches and cells without an “X” indicate no mismatch observed