Understanding star formation rates (SFR) is a central goal of modern star-formation models, which mainly involve gravity, turbulence and, in some cases, magnetic fields (Bfields) 1,2 . However, a connection between B-fields and SFR has never been observed. Here, a comparison between the surveys of SFR 3,4 and a study of cloud-field alignment 5 -which revealed a bimodal (parallel or perpendicular) alignment-shows consistently lower SFR per solar mass for clouds almost perpendicular to the B-fields. This is evidence of B-fields being a primary regulator of SFR. The perpendicular alignment possesses a significantly higher magnetic flux than the parallel alignment and thus a stronger support of the gas against selfgravity. This results in overall lower masses of the fragmented components, which are in agreement with the lower SFR.It is not difficult to imagine that the SFR of a molecular cloud should be related to the mass of gas it contains, due to self-gravity. On the whole, this trend has indeed been observed 3,4 . On the other hand, star formation efficiencies of molecular clouds are usually just a few percent 6 while cloud ages are at least comparable to their free-fall time (~Myr) 7 . This requires other forces to slow down gravitational contraction 8 . In addition, clouds with a similar mass and age can have different SFR; for example, it is well known that the Ophiuchus cloud has a significantly higher SFR than its neighbour, the Pipe Nebula 3,4 . In Figure 1, we can see a significant difference between the two dark clouds: Ophiuchus is accompanied by the colourful nebulae, which is a signature of active stellar feedback, while Pipe looks very quiescent in comparison. No one knows the reason for these differences in SFR. Another good example is the pair of Rosette and G216-2.5 molecular clouds 9 , where turbulence had been suspected as the cause of their very different SFRs. But later their turbulent velocity spectra were found almost identical 9 . Recently, a good agreement has been discovered between the empirical column density threshold for cloud contraction and the magnetic critical column density of the Galactic field (~10 µG) 5,10 . For densities lower than this threshold, the field strength is independent of densities 11 ; i.e., gas must accumulate along field lines. Also, a bimodal cloud-field alignment, which is another signature of field-regulated (sub-Alfvenic turbulence) cloud formation, was recently observed in the Gould Belt 5 , where, interestingly, Pipe and Ophiuchus (Figure 1) are aligned differently from their local fields. More intriguingly, it is the one aligned with the B-field that holds the higher SFR. Together with another observed piece of evidence that Galactic B-field direction anchors deeply into cloud cores 12 and thus plays a role in cloud fragmentation 13 , we were prompted to survey whether the cloud-field alignment has a connection with SFR.Luckily, both the cloud-field alignment 5 and SFR 3,4 of the Gould Belt clouds have been very well studied (Table 1). Both Heiderman e...

The Zeeman effect has been the only method to directly probe the magnetic field strength in molecular clouds. The Bayesian analysis of Zeeman measurements carried out by Crutcher et al. is the only reference for cloud magnetic field strength. Here we extended their model and Bayesian analysis of the relation between field strength (B) and volume density (n) in the following three directions based on the recent observational and theoretical development. First, we take R, the observational uncertainty of n, as a parameter to be estimated from data. Second, the restriction of α, the index of the B–n relationship, is relieved from [0, 0.75] to [0, 1]. Third, we allow f, the minimum-to-maximum B ratio, to vary with n. Our results show that taking R as a parameter provides a better fitting to the B–n relationship and much more reliable estimates on R, f, and the changing point of α. Arguably our most important finding is that α cannot be reliably estimated by any of the models studied here, either from us or Crutcher et al., if R > 2, which is indeed the case from our estimate. This is the so-called errors-in-variables bias, a well known problem for statisticians.

N6-methyladenine (6mA) is an important DNA modification form associated with a wide range of biological processes. Identifying accurately 6mA sites on a genomic scale is crucial for under-standing of 6mA’s biological functions. However, the existing experimental techniques for detecting 6mA sites are cost-ineffective, which implies the great need of developing new computational methods for this problem. In this paper, we developed, without requiring any prior knowledge of 6mA and manually crafted sequence features, a deep learning framework named Deep6mA to identify DNA 6mA sites, and its performance is superior to other DNA 6mA prediction tools. Specifically, the 5-fold cross-validation on a benchmark dataset of rice gives the sensitivity and specificity of Deep6mA as 92.96% and 95.06%, respectively, and the overall prediction accuracy is 94%. Importantly, we find that the sequences with 6mA sites share similar patterns across different species. The model trained with rice data predicts well the 6mA sites of other three species: Arabidopsis thaliana, Fragaria vesca and Rosa chinensis with a prediction accuracy over 90%. In addition, we find that (1) 6mA tends to occur at GAGG motifs, which means the sequence near the 6mA site may be conservative; (2) 6mA is enriched in the TATA box of the promoter, which may be the main source of its regulating downstream gene expression.

No abstract