Search for: All records

Gradual typing has emerged as a promising typing discipline for reconciling static and dynamic typing, which have respective strengths and shortcomings. Thanks to its promises, gradual typing has gained tremendous momentum in both industry and academia. A main challenge in gradual typing is that, however, the performance of its programs can often be unpredictable, and adding or removing the type of a a single parameter may lead to wild performance swings. Many approaches have been proposed to optimize gradual typing performance, but little work has been done to aid the understanding of the performance landscape of gradual typing and navigating the migration process (which adds type annotations to make programs more static) to avert performance slowdowns. Motivated by this situation, this work develops a machine-learning-based approach to predict the performance of each possible way of adding type annotations to a program. On top of that, many supports for program migrations could be developed, such as finding the most performant neighbor of any given configuration. Our approach gauges runtime overheads of dynamic type checks inserted by gradual typing and uses that information to train a machine learning model, which is used to predict the running time of gradual programs. We have evaluated our approach on 12 Python benchmarks for both guarded and transient semantics. For guarded semantics, our evaluation results indicate that with only 40 training instances generated from each benchmark, the predicted times for all other instances differ on average by 4% from the measured times. For transient semantics, the time difference ratio is higher but the time difference is often within 0.1 seconds. 

Confocal Raman microscopy was applied to detect structural change within individual particles of low-density polyethylene (LDPE) following chemical and electrochemical processing steps that aimed to facilitate material decomposition. A high numerical aperture (NA) oil-immersion objective enabled depth-profiling through the near surface region (20 μm–40 μm) of irregularly shaped particles with an axial spatial resolution < 2 μm estimated from measurements of instrument detection efficiency profiles. Changes in vibrational bands sensitive to polyethylene crystallinity were evident following treatments and linked to the release of low molecular weight compounds present as additives and products of processing. Effects of processing were probed by monitoring the rise of Raman scattering intensity in vibrational modes associated with polyethylene chains in a zig-zag (trans) conformation near 1128 cm^–1, 1294 cm^–1, and 1418 cm^–1, signaling chain clustering and development of organized, crystalline-like assemblies. Pristine LDPE particles displayed a uniform structure across the near surface region, while particles treated initially with chemical extractant and then further processed displayed increasingly enhanced crystallinity up to the maximum depth probed (40 μm). As a step toward measurements on ensembles of particles, least squares modeling was adapted to derive pure component spectra reflecting crystallinity change within spectral datasets. The work demonstrates high spatial resolution Raman depth-profiling for the characterization of processed polymers using a high NA immersion objective to overcome the limitations of air-objectives often used for confocal Raman microscopy. 

Membraneless organelles are RNA–protein assemblies which have been implicated in post‐transcriptional control. Germ cells form membraneless organelles referred to as germ granules, which contain conserved proteins including Tudor domain‐containing scaffold polypeptides and their partner proteins that interact with Tudor domains. Here, we show that inDrosophila, different germ granule proteins associate with the multi‐domain Tudor protein using different numbers of Tudor domains. Furthermore, these proteins compete for interaction with Tudorin vitroand, surprisingly, partition to distinct and poorly overlapping clusters in germ granulesin vivo. This partition results in minimization of the competition. Our data suggest that Tudor forms structurally different configurations with different partner proteins which dictate different biophysical properties and phase separation parameters within the same granule. 

The assembly of large RNA-protein granules occurs in germ cells of many animals and these germ granules have provided a paradigm to study structure-functional aspects of similar structures in different cells. Germ granules in Drosophila oocyte’s posterior pole (polar granules) are composed of RNA, in the form of homotypic clusters, and proteins required for germline development. In the granules, Piwi protein Aubergine binds to a scaffold protein Tudor, which contains 11 Tudor domains. Using a super-resolution microscopy, we show that surprisingly, Aubergine and Tudor form distinct clusters within the same polar granules in early Drosophila embryos. These clusters partially overlap and, after germ cells form, they transition into spherical granules with the structural organization unexpected from these interacting proteins: Aubergine shell around the Tudor core. Consistent with the formation of distinct clusters, we show that Aubergine forms homo-oligomers and using all purified Tudor domains, we demonstrate that multiple domains, distributed along the entire Tudor structure, interact with Aubergine. Our data suggest that in polar granules, Aubergine and Tudor are assembled into distinct phases, partially mixed at their “interaction hubs”, and that association of distinct protein clusters may be an evolutionarily conserved mechanism for the assembly of germ granules.