Expedient production of site specifically nucleobase-labelled or hypermodified RNA with engineered thermophilic DNA polymerases.

Innovative approaches to controlled nucleobase-modified RNA synthesis are urgently needed to support RNA biology exploration and to synthesize potential RNA therapeutics. Here we present a strategy for enzymatic construction of nucleobase-modified RNA based on primer-dependent engineered thermophilic DNA polymerases - SFM4-3 and TGK. We demonstrate introduction of one or several different base-modified nucleotides in one strand including hypermodified RNA containing all four modified nucleotides bearing four different substituents, as well as strategy for primer segment removal. We also show facile site-specific or segmented introduction of fluorophores or other functional groups at defined positions in variety of RNA molecules, including structured or long mRNA. Intriguing translation efficacy of single-site modified mRNAs underscores the necessity to study isolated modifications placed at designer positions to disentangle their biological effects and enable development of improved mRNA therapeutics. Our toolbox paves the way for more precise dissecting RNA structures and functions, as well as for construction of diverse types of base-functionalized RNA for therapeutic applications and diagnostics.

Design, Synthesis, and Biochemical and Biological Evaluation of Novel 7-Deazapurine Cyclic Dinucleotide Analogues as STING Receptor Agonists.

Cyclic dinucleotides (CDNs) are second messengers that activate stimulator of interferon genes (STING). The cGAS-STING pathway plays a promising role in cancer immunotherapy. Here, we describe the synthesis of CDNs containing 7-substituted 7-deazapurine moiety. We used mouse cyclic GMP-AMP synthase and bacterial dinucleotide synthases for the enzymatic synthesis of CDNs. Alternatively, 7-(het)aryl 7-deazapurine CDNs were prepared by Suzuki-Miyaura cross-couplings. New CDNs were tested in biochemical and cell-based assays for their affinity to human STING. Eight CDNs showed better activity than 2'3'-cGAMP, the natural ligand of STING. The effect on cytokine and chemokine induction was also evaluated. The best activities were observed for CDNs bearing large aromatic substituents that point above the CDN molecule. We solved four X-ray structures of complexes of new CDNs with human STING. We observed π-π stacking interactions between the aromatic substituents and Tyr240 that are involved in the stabilization of CDN-STING complexes.

Supervised learning with quantum-enhanced feature spaces.

Machine learning and quantum computing are two technologies that each have the potential to alter how computation is performed to address previously untenable problems. Kernel methods for machine learning are ubiquitous in pattern recognition, with support vector machines (SVMs) being the best known method for classification problems. However, there are limitations to the successful solution to such classification problems when the feature space becomes large, and the kernel functions become computationally expensive to estimate. A core element in the computational speed-ups enabled by quantum algorithms is the exploitation of an exponentially large quantum state space through controllable entanglement and interference. Here we propose and experimentally implement two quantum algorithms on a superconducting processor. A key component in both methods is the use of the quantum state space as feature space. The use of a quantum-enhanced feature space that is only efficiently accessible on a quantum computer provides a possible path to quantum advantage. The algorithms solve a problem of supervised learning: the construction of a classifier. One method, the quantum variational classifier, uses a variational quantum circuit1,2 to classify the data in a way similar to the method of conventional SVMs. The other method, a quantum kernel estimator, estimates the kernel function on the quantum computer and optimizes a classical SVM. The two methods provide tools for exploring the applications of noisy intermediate-scale quantum computers3 to machine learning.

Quantum Schur Sampling Circuits can be Strongly Simulated.

Permutational quantum computing (PQC) [Quantum Inf. Comput., 10, 470-497 (2010)QICUAW1533-7146] is a natural quantum computational model conjectured to capture nonclassical aspects of quantum computation. An argument backing this conjecture was the observation that there was no efficient classical algorithm for estimation of matrix elements of the S_{n} irreducible representation matrices in the Young's orthogonal form, which correspond to transition amplitudes of a broad class of PQC circuits. This problem can be solved with a PQC machine in polynomial time, but no efficient classical algorithm for the problem was previously known. Here we give a classical algorithm that efficiently approximates the transition amplitudes up to polynomial additive precision and hence solves this problem. We further extend our discussion to show that transition amplitudes of a broader class of quantum circuits-the quantum Schur sampling circuits-can also be efficiently classically approximated.