Transfer learning for cross-context prediction of protein expression from 5'UTR sequence.

Model-guided DNA sequence design can accelerate the reprogramming of living cells. It allows us to engineer more complex biological systems by removing the need to physically assemble and test each potential design. While mechanistic models of gene expression have seen some success in supporting this goal, data-centric, deep learning-based approaches often provide more accurate predictions. This accuracy, however, comes at a cost - a lack of generalization across genetic and experimental contexts that has limited their wider use outside the context in which they were trained. Here, we address this issue by demonstrating how a simple transfer learning procedure can effectively tune a pre-trained deep learning model to predict protein translation rate from 5' untranslated region (5'UTR) sequence for diverse contexts in Escherichia coli using a small number of new measurements. This allows for important model features learnt from expensive massively parallel reporter assays to be easily transferred to new settings. By releasing our trained deep learning model and complementary calibration procedure, this study acts as a starting point for continually refined model-based sequence design that builds on previous knowledge and future experimental efforts.

Effective design and inference for cell sorting and sequencing based massively parallel reporter assays.

MOTIVATION: The ability to measure the phenotype of millions of different genetic designs using Massively Parallel Reporter Assays (MPRAs) has revolutionized our understanding of genotype-to-phenotype relationships and opened avenues for data-centric approaches to biological design. However, our knowledge of how best to design these costly experiments and the effect that our choices have on the quality of the data produced is lacking. RESULTS: In this article, we tackle the issues of data quality and experimental design by developing FORECAST, a Python package that supports the accurate simulation of cell-sorting and sequencing-based MPRAs and robust maximum likelihood-based inference of genetic design function from MPRA data. We use FORECAST's capabilities to reveal rules for MPRA experimental design that help ensure accurate genotype-to-phenotype links and show how the simulation of MPRA experiments can help us better understand the limits of prediction accuracy when this data are used for training deep learning-based classifiers. As the scale and scope of MPRAs grows, tools like FORECAST will help ensure we make informed decisions during their development and the most of the data produced. AVAILABILITY AND IMPLEMENTATION: The FORECAST package is available at: https://gitlab.com/Pierre-Aurelien/forecast. Code for the deep learning analysis performed in this study is available at: https://gitlab.com/Pierre-Aurelien/rebeca.

Design and Analysis of Massively Parallel Reporter Assays Using FORECAST.

Machine learning is revolutionizing molecular biology and bioengineering by providing powerful insights and predictions. Massively parallel reporter assays (MPRAs) have emerged as a particularly valuable class of high-throughput technique to support such algorithms. MPRAs enable the simultaneous characterization of thousands or even millions of genetic constructs and provide the large amounts of data needed to train models. However, while the scale of this approach is impressive, the design of effective MPRA experiments is challenging due to the many factors that can be varied and the difficulty in predicting how these will impact the quality and quantity of data obtained. Here, we present a computational tool called FORECAST, which can simulate MPRA experiments based on fluorescence-activated cell sorting and subsequent sequencing (commonly referred to as Flow-seq or Sort-seq experiments), as well as carry out rigorous statistical estimation of construct performance from this type of experimental data. FORECAST can be used to develop workflows to aid the design of MPRA experiments and reanalyze existing MPRA data sets.

Sequencing enabling design and learning in synthetic biology.

The ability to read and quantify nucleic acids such as DNA and RNA using sequencing technologies has revolutionized our understanding of life. With the emergence of synthetic biology, these tools are now being put to work in new ways - enabling de novo biological design. Here, we show how sequencing is supporting the creation of a new wave of biological parts and systems, as well as providing the vast data sets needed for the machine learning of design rules for predictive bioengineering. However, we believe this is only the tip of the iceberg and end by providing an outlook on recent advances that will likely broaden the role of sequencing in synthetic biology and its deployment in real-world environments.

Magnetic and luminescent coordination networks based on imidazolium salts and lanthanides for sensitive ratiometric thermometry.

The synthesis and characterization of six new lanthanide networks [Ln(L)(ox)(H2O)] with Ln = Eu3+, Gd3+, Tb3+, Dy3+, Ho3+ and Yb3+ is reported. They were synthesized by solvo-ionothermal reaction of lanthanide nitrate Ln(NO3)3·xH2O with the 1,3-bis(carboxymethyl)imidazolium [HL] ligand and oxalic acid (H2ox) in a water/ethanol solution. The crystal structure of these compounds has been solved on single crystals and the magnetic and luminescent properties have been investigated relying on intrinsic properties of the lanthanide ions. The synthetic strategy has been extended to mixed lanthanide networks leading to four isostructural networks of formula [Tb1- x Eu x (L)(ox)(H2O)] with x = 0.01, 0.03, 0.05 and 0.10. These materials were assessed as luminescent ratiometric thermometers based on the emission intensities of ligand, Tb3+ and Eu3+. The best sensitivities were obtained using the ratio between the emission intensities of Eu3+ (5D0→7F2 transition) and of the ligand as the thermometric parameter. [Tb0.97Eu0.03(L)(ox)(H2O)] was found to be one of the best thermometers among lanthanide-bearing coordination polymers and metal-organic frameworks, operative in the physiological range with a maximum sensitivity of 1.38%·K-1 at 340 K.

Giant light-harvesting nanoantenna for single-molecule detection in ambient light.

Here, we explore the enhancement of single molecule emission by polymeric nano-antenna that can harvest energy from thousands of donor dyes to a single acceptor. In this nano-antenna, the cationic dyes are brought together in very close proximity using bulky counterions, thus enabling ultrafast diffusion of excitation energy (≤30 fs) with minimal losses. Our 60-nm nanoparticles containing >10,000 rhodamine-based donor dyes can efficiently transfer energy to 1-2 acceptors resulting in an antenna effect of ~1,000. Therefore, single Cy5-based acceptors become 25-fold brighter than quantum dots QD655. This unprecedented amplification of the acceptor dye emission enables observation of single molecules at illumination powers (1-10 mW cm-2) that are >10,000-fold lower than typically required in single-molecule measurements. Finally, using a basic setup, which includes a 20X air objective and a sCMOS camera, we could detect single Cy5 molecules by simply shining divergent light on the sample at powers equivalent to sunlight.

Photon management properties of rare-earth (Nd,Yb,Sm)-doped CeO2 films prepared by pulsed laser deposition.

CeO2 is a promising material for applications in optoelectronics and photovoltaics due to its large band gap and values of the refractive index and lattice parameters, which are suitable for silicon-based devices. In this study, we show that trivalent Sm, Nd and Yb ions can be successfully inserted and optically activated in CeO2 films grown at a relatively low deposition temperature (400 °C), which is compatible with inorganic photovoltaics. CeO2 thin films can therefore be efficiently functionalized with photon-management properties by doping with trivalent rare earth (RE) ions. Structural and optical analyses provide details of the electronic level structure of the films and of their energy transfer mechanisms. In particular, we give evidence of the existence of an absorption band centered at 350 nm from which energy transfer to rare earth ions occurs. The transfer mechanisms can be completely explained only by considering the spontaneous migration of Ce(3+) ions in CeO2 at a short distance from the RE(3+) ions. The strong absorption cross section of the f-d transitions in Ce(3+) ions efficiently intercepts the UV photons of the solar spectrum and therefore strongly increases the potential of these layers as downshifters and downconverters.

Distance dependence of the energy transfer rate from a single semiconductor nanostructure to graphene.

The near-field Coulomb interaction between a nanoemitter and a graphene monolayer results in strong Förster-type resonant energy transfer and subsequent fluorescence quenching. Here, we investigate the distance dependence of the energy transfer rate from individual, (i) zero-dimensional CdSe/CdS nanocrystals and (ii) two-dimensional CdSe/CdS/ZnS nanoplatelets to a graphene monolayer. For increasing distances d, the energy transfer rate from individual nanocrystals to graphene decays as 1/d(4). In contrast, the distance dependence of the energy transfer rate from a two-dimensional nanoplatelet to graphene deviates from a simple power law but is well described by a theoretical model, which considers a thermal distribution of free excitons in a two-dimensional quantum well. Our results show that accurate distance measurements can be performed at the single particle level using graphene-based molecular rulers and that energy transfer allows probing dimensionality effects at the nanoscale.

A complex study of the fast blue luminescence of oxidized silicon nanocrystals: the role of the core.

Silicon nanocrystals (SiNCs) smaller than 5 nm are a material with strong visible photoluminescence (PL). However, the physical origin of the PL, which, in the case of oxide-passivated SiNCs, is typically composed of a slow-decaying red-orange band (S-band) and of a fast-decaying blue-green band (F-band), is still not fully understood. Here we present a physical interpretation of the F-band origin based on the results of an experimental study, in which we combine temperature (4-296 K), temporally (picosecond resolution) and spectrally resolved luminescence spectroscopy of free-standing oxide-passivated SiNCs. Our complex study shows that the F-band red-shifts only by 35 meV with increasing temperature, which is almost 6 times less than the red-shift of the S-band in a similar temperature range. In addition, the F-band characteristic decay time obtained from a stretched-exponential fit decreases only slightly with increasing temperature. These data strongly suggest that the F-band arises from the core-related quasi-direct radiative recombination governed by slowly thermalizing photoholes.

Effect of ball-milling and Fe-/Al-doping on the structural aspect and visible light photocatalytic activity of TiO2 towards Escherichia coli bacteria abatement.

Escherichia coli abatement was studied in liquid phase under visible light in the presence of two commercial titania photocatalysts, and of Fe- and Al-doped titania samples prepared by high energy ball-milling. The two commercial titania photocatalysts, Aeroxide P25 (Evonik industries) exhibiting both rutile and anatase structures and MPT625 (Ishihara Sangyo Kaisha), a Fe-, Al-, P- and S-doped titania exhibiting only the rutile phase, are active suggesting that neither the structure nor the doping is the driving parameter. Although the MPT625 UV-visible spectrum is shifted towards the visible domain with respect to the P25 one, the effect on bacteria is not increased. On the other hand, the ball milled iron-doped P25 samples exhibit low activities in bacteria abatement under visible light due to charge recombinations unfavorable to catalysis as shown by photoluminescence measurements. While doping elements are in interstitial positions within the rutile structure in MPT625 sample, they are located at the surface in ball milled samples and in isolated octahedral units according to (57)Fe Mössbauer spectrometry. The location of doping elements at the surface is suggested to be responsible for the sample cytotoxicity observed in the dark.

All-optical trion generation in single-walled carbon nanotubes.

We present evidence of all-optical trion generation and emission in pristine single-walled carbon nanotubes (SWCNTs). Luminescence spectra, recorded on individual SWCNTs over a large cw excitation intensity range, show trion emission peaks redshifted with respect to the bright exciton peak. Clear chirality dependence is observed for 22 separate SWCNT species, allowing for determination of electron-hole exchange interaction and trion binding energy contributions. Luminescence data together with ultrafast pump-probe experiments on chirality-sorted bulk samples suggest that exciton-exciton annihilation processes generate dissociated carriers that allow for trion creation upon a subsequent photon absorption event.

Correlation of structural properties with energy transfer of Eu-doped ZnO thin films prepared by sol-gel process and magnetron reactive sputtering.

We investigated the structural and optical properties of Eu-doped ZnO thin films made by sol-gel technique and magnetron reactive sputtering on Si (100) substrate. The films elaborated by sol-gel process are polycrystalline while the films made by sputtering show a strongly textured growth along the c-axis. X-ray diffraction patterns and transmission electron microscopy analysis show that all samples are free of spurious phases. The presence of Eu(2+) and Eu(3+) into the ZnO matrix has been confirmed by x-ray photoemission spectroscopy. This means that a small fraction of Europium substitutes Zn(2+) as Eu(2+) into the ZnO matrix; the rest of Eu being in the trivalent state. This is probably due to the formation of Eu(2)O(3) oxide at the surface of ZnO particles. This is at the origin of the strong photoluminescence band observed at 2 eV, which is characteristic of the (5)D(0)-->(7)F(2) Eu(3+) transition. In addition the photoluminescence excitonic spectra showed efficient energy transfer from the ZnO matrix to the Eu(3+) ion, which is qualitatively similar for both films although the sputtered films have a better structural quality compared to the sol-gel process grown films.