Table 1.

Re-engineering of riboswitches

Target				RNAinverse		RNAensign
						100K		310K		1000K
	length		G		G		G		G		G
Ade ydhL gene OFF	110	0.238	−17.2	0.118	−25.0	0.169	−25.5	0.100	−25.7	0.116	−24.7
Ade ydhL gene ON	110	0.238	17.2	0.237	−6.1	0.343	−9.2	0.352	−3.7
Ade add gene OFF	113	0.446	−1.8	0.253	−5.8	0.075	−9.2	0.073	−15.2	0.088	−8.8
Ade add gene ON	113	0.446	1.8	0.307	−9.8			0.159	−9.8	0.189	−10.8
c-di-GMP OFF	124	0.381	−8.8	0.237	−25.3	0.233	−27.7	0.225	−38.0	0.304	−27.3
c-di-GMP ON	124	0.381	8.8	0.292	−15.1
SAM OFF	134	0.302	−15.3	0.123	−15.2	0.213	−29.0	0.183	−22.7	0.147	−22.5
SAM ON	134	0.302	15.3
xpt-pubX OFF	148	0.073	−18.3	0.101	−22.7			0.116	−31.6	0.117	−18.9
xpt-pubX ON	148	0.073	18.3	0.435	−5.8

Re-engineering of riboswitches using RNAinverse (we report the best results among 1000 runs) and RNA-ensign. Only solutions with less than 10% of mutations have been considered. We selected the solution with the lowest entropy. For each method, we report the entropy and the difference of energies G between the two conformation ‘ON’ and ‘OFF’ of the riboswitches. To illustrate the versatility of our approach, we ran RNA-ensign with three different formal temperatures: T = 100 (increased thermodynamic pressure), T = 310 (default) and T = 1000 (reduced thermodynamic pressure).