User:ElNando888/Blog/Truth tables
< User:ElNando888 | Blog
In the latest Eterna Open Group Meeting, Rhiju mentioned among many other things, that one of the future goals of the Das Lab would be to build "computing bio-machines". And he went on talking about truth tables, and mentioning that some labs around the world were currently working on designing bio-molecules that would initiate apostosis (cell death) upon detection of say, 4 different RNA "signals".
I've been trying to picture what was meant. So here is an example I came up with, which should be regarded as a thought experiment.
Let's assume we want to build something that would detect the simultaneous presence of 2 different miRNAs.
- mir-1 = CACACUCAUGUGGCCUCCAGCA
- mir-2 = UUCCCAUUCCACAAGCCCUGAU
And we want the signal to be the MS2 hairpin.
A simple way to detect a given RNA sequence would be to use its complementary sequence, so as to allow for the formation of a dimer. And to detect 2 of them, we need to have both complements.
UGCUGGAGGCCACAUGAGUGUG-UUCCGCAUGA-ACAUGAGGAUGACCCAUGU-GGUGCACCUC-AUCAGGGCUUGUGGAAUGGGAA
The above sequence does just that. The complement of mir-1 is at the 5' end, and the complement of mir-2 is at the 3' one. In between, we have the MS2 sequence, and 2 buffers of nucleotides carefully chosen for the specific purpose of creating the desired 'truth table'.
- mir-1 absent (0), mir-2 absent (0)
MS2 hairpin not formed (0) - mir-1 absent (0), mir-2 present (1)
MS2 hairpin not formed (0) - mir-1 present (1), mir-2 absent (0)
MS2 hairpin not formed (0) - mir-1 present (1), mir-2 present (1)
MS2 hairpin formed (1)
(Note: all the above MFE predictions are from Vienna 2.1.x)
And this would represent a 'AND' truth table, in boolean logic.
Now, I'm quite uncertain how close or far this is from what Rhiju is envisioning.
And even if I'm actually getting the right idea, I'm having a hard time picturing how to factor 2 more signals in such constructs. Anyone got brilliant ideas?
