Revisiting Consistency for Semi-Supervised Semantic Segmentation.

Semi-supervised learning is an attractive technique in practical deployments of deep models since it relaxes the dependence on labeled data. It is especially important in the scope of dense prediction because pixel-level annotation requires substantial effort. This paper considers semi-supervised algorithms that enforce consistent predictions over perturbed unlabeled inputs. We study the advantages of perturbing only one of the two model instances and preventing the backward pass through the unperturbed instance. We also propose a competitive perturbation model as a composition of geometric warp and photometric jittering. We experiment with efficient models due to their importance for real-time and low-power applications. Our experiments show clear advantages of (1) one-way consistency, (2) perturbing only the student branch, and (3) strong photometric and geometric perturbations. Our perturbation model outperforms recent work and most of the contribution comes from the photometric component. Experiments with additional data from the large coarsely annotated subset of Cityscapes suggest that semi-supervised training can outperform supervised training with coarse labels. Our source code is available at https://github.com/Ivan1248/semisup-seg-efficient.

Amino acid (acyl carrier protein) ligase-associated biosynthetic gene clusters reveal unexplored biosynthetic potential.

This study aimed to evaluate the postulated cellular function of a novel family of amino acid (acyl carrier protein) ligases (AALs) in natural product biosynthesis. Here, we analyzed the manually curated, putative, aal-associated natural product biosynthetic gene clusters (NP BGCs) using two computational platforms for NP prediction, antiSMASH-BiG-SCAPE-CORASON and DeepBGC. The detected BGCs included a diversity of type I polyketide/nonribosomal peptide (PKS/NRPS) hybrid BGCs, exemplified by the guadinomine BGC, which suggested a dedicated function of AALs in the biosynthesis of rare (2S)-aminomalonyl-ACP extension units. Besides modular PKS/NRPSs and NRPSs, AAL-associated BGCs were predicted to assemble arylpolyenes, ladderane lipids, phosphonates, aminoglycosides, ß-lactones, and thioamides of both nonribosomal and ribosomal origins. Additionally, we revealed a frequent association of AALs with putative, seldom observed transglutaminase-like and BtrH-like transferases of the cysteine protease superfamily, which may form larger families of ACP-dependent amide bond catalysts used in NP synthesis. Our results disclosed an exceptional chemical novelty and biosynthetic potential of the AAL-associated BGCs in NP biosynthesis. The presented in silico evidence supports the initial hypothesis and provides an important foundation for future experimental studies aimed at discovering novel pharmaceutically relevant active compounds.

Evolution of Functionally Enhanced α-l-Threofuranosyl Nucleic Acid Aptamers.

Synthetic genetic polymers (xeno-nucleic acids, XNAs) have the potential to transition aptamers from laboratory tools to therapeutic agents, but additional functionality is needed to compete with antibodies. Here, we describe the evolution of a biologically stable artificial genetic system composed of α-l-threofuranosyl nucleic acid (TNA) that facilitates the production of backbone- and base-modified aptamers termed "threomers" that function as high quality protein capture reagents. Threomers were discovered against two prototypical protein targets implicated in human diseases through a combination of in vitro selection and next-generation sequencing using uracil nucleotides that are uniformly equipped with aromatic side chains commonly found in the paratope of antibody-antigen crystal structures. Kinetic measurements reveal that the side chain modifications are critical for generating threomers with slow off-rate binding kinetics. These findings expand the chemical space of evolvable non-natural genetic systems to include functional groups that enhance protein target binding by mimicking the structural properties of traditional antibodies.

COMPUTER BASED EYE TRACKER FOR DETECTION OF MANIFEST STRABISMUS.

Strabismus is a common disorder in which eyes are not aligned with their optical axis, resulting in an abnormal binocular interaction, thus leading to amblyopia. Accordingly, early detection and diagnosis are mandatory. Despite technology development and constant knowledge growth, the foundations of physiology and the diagnosis of strabismus were set back in the 19th or early 20th century and have not changed since. In this paper, a novel, properly tested and evaluated eye-tracking based method for manifest strabismus diagnosis is presented. The evaluation showed the aforementioned method to have both high sensitivity and specificity in detecting manifest strabismus without the need for a skilled examiner, thus being suitable for population screening and highlighting the need for future research regarding testing of latent strabismus.

Functional and mechanistic studies of XPC DNA-repair complex as transcriptional coactivator in embryonic stem cells.

The embryonic stem cell (ESC) state is transcriptionally controlled by OCT4, SOX2, and NANOG with cofactors, chromatin regulators, noncoding RNAs, and other effectors of signaling pathways. Uncovering components of these regulatory circuits and their interplay provides the knowledge base to deploy ESCs and induced pluripotent stem cells. We recently identified the DNA-repair complex xeroderma pigmentosum C (XPC)-RAD23B-CETN2 as a stem cell coactivator (SCC) required for OCT4/SOX2 transcriptional activation. Here we investigate the role of SCC genome-wide in murine ESCs by mapping regions bound by RAD23B and analyzing transcriptional profiles of SCC-depleted ESCs. We establish OCT4 and SOX2 as the primary transcription factors recruiting SCC to regulatory regions of pluripotency genes and identify the XPC subunit as essential for interaction with the two proteins. The present study reveals new mechanistic and functional aspects of SCC transcriptional activity, and thus underscores the diversified functions of this regulatory complex.

TAF7L modulates brown adipose tissue formation.

Brown adipose tissue (BAT) plays an essential role in metabolic homeostasis by dissipating energy via thermogenesis through uncoupling protein 1 (UCP1). Previously, we reported that the TATA-binding protein associated factor 7L (TAF7L) is an important regulator of white adipose tissue (WAT) differentiation. In this study, we show that TAF7L also serves as a molecular switch between brown fat and muscle lineages in vivo and in vitro. In adipose tissue, TAF7L-containing TFIID complexes associate with PPARγ to mediate DNA looping between distal enhancers and core promoter elements. Our findings suggest that the presence of the tissue-specific TAF7L subunit in TFIID functions to promote long-range chromatin interactions during BAT lineage specification.

Taf7l cooperates with Trf2 to regulate spermiogenesis.

TATA-binding protein (TBP)-associated factor 7l (Taf7l; a paralogue of Taf7) and TBP-related factor 2 (Trf2) are components of the core promoter complex required for gene/tissue-specific transcription of protein-coding genes by RNA polymerase II. Previous studies reported that Taf7l knockout (KO) mice exhibit structurally abnormal sperm, reduced sperm count, weakened motility, and compromised fertility. Here we find that continued backcrossing of Taf7l(-/Y) mice from N5 to N9 produced KO males that are essentially sterile. Genome-wide expression profiling by mRNA-sequencing analysis of wild-type (WT) and Taf7l(-/Y) (KO) testes revealed that Taf7l ablation impairs the expression of many postmeiotic spermatogenic-specific as well as metabolic genes. Importantly, histological analysis of testes revealed that Taf7l(-/Y) mice develop postmeiotic arrest at the first stage of spermiogenesis, phenotypically similar to Trf2(-/-) mice, but distinct from Taf4b(-/-) mice. Indeed, we find that Taf7l and Trf2 coregulate postmeiotic genes, but none of Taf4b-regulated germ stem cell genes in testes. Genome-wide ChIP-sequencing studies indicate that TAF7L binds to promoters of activated postmeiotic genes in testis. Moreover, biochemical studies show that TAF7L associates with TRF2 both in vitro and in testis, suggesting that TAF7L likely cooperates directly with TRF2 at promoters of a subset of postmeiotic genes to regulate spermiogenesis. Our findings thus provide a previously undescribed mechanism for cell-type-specific transcriptional control involving an interaction between a "nonprototypic" core promoter recognition factor (Trf2) and an orphan TAF subunit (Taf7l) in mammalian testis-specific gene transcription.

Dual functions of TAF7L in adipocyte differentiation.

The diverse transcriptional mechanisms governing cellular differentiation and development of mammalian tissue remains poorly understood. Here we report that TAF7L, a paralogue of TFIID subunit TAF7, is enriched in adipocytes and white fat tissue (WAT) in mouse. Depletion of TAF7L reduced adipocyte-specific gene expression, compromised adipocyte differentiation, and WAT development as well. Ectopic expression of TAF7L in myoblasts reprograms these muscle precursors into adipocytes upon induction. Genome-wide mRNA-seq expression profiling and ChIP-seq binding studies confirmed that TAF7L is required for activating adipocyte-specific genes via a dual mechanism wherein it interacts with PPARγ at enhancers and TBP/Pol II at core promoters. In vitro binding studies confirmed that TAF7L forms complexes with both TBP and PPARγ. These findings suggest that TAF7L plays an integral role in adipocyte gene expression by targeting enhancers as a cofactor for PPARγ and promoters as a component of the core transcriptional machinery.DOI:http://dx.doi.org/10.7554/eLife.00170.001.

A DNA repair complex functions as an Oct4/Sox2 coactivator in embryonic stem cells.

The transcriptional activators Oct4, Sox2, and Nanog cooperate with a wide array of cofactors to orchestrate an embryonic stem (ES) cell-specific gene expression program that forms the molecular basis of pluripotency. Here, we report using an unbiased in vitro transcription-biochemical complementation assay to discover a multisubunit stem cell coactivator complex (SCC) that is selectively required for the synergistic activation of the Nanog gene by Oct4 and Sox2. Purification, identification, and reconstitution of SCC revealed this coactivator to be the trimeric XPC-nucleotide excision repair complex. SCC interacts directly with Oct4 and Sox2 and is recruited to the Nanog and Oct4 promoters as well as a majority of genomic regions that are occupied by Oct4 and Sox2. Depletion of SCC/XPC compromised both pluripotency in ES cells and somatic cell reprogramming of fibroblasts to induced pluripotent stem (iPS) cells. This study identifies a transcriptional coactivator with diversified functions in maintaining ES cell pluripotency and safeguarding genome integrity.

Biased coarse-grained molecular dynamics simulation approach for flexible fitting of X-ray structure into cryo electron microscopy maps.

Several approaches have been introduced to interpret, in terms of high-resolution structure, low-resolution structural data as obtained from cryo-EM. As conformational changes are often observed in biological molecules, these techniques need to take into account the flexibility of proteins. Flexibility has been described in terms of movement between rigid domains and between rigid secondary structure elements, which present some limitations for studying dynamical properties. Normal mode analysis has also been used, but is limited to medium resolution data. All-atom molecular dynamics fitting techniques are more appropriate to fit structures into higher-resolution data as full protein flexibility is considered, but are cumbersome in terms of computational time. Here, we introduce a coarse-grained approach; a Go-model was used to represent biological molecules, combined with biased molecular dynamics to reproduce accurately conformational transitions. Illustrative examples on simulated data are shown. Accurate fittings can be obtained for resolution ranging from 5 to 20A. The approach was also tested on experimental data of Elongation Factor G and Escherichia coli RNA polymerase, where its validity is compared to previous models obtained from different techniques. This comparison demonstrates that quantitative flexible techniques, as opposed to manual docking, need to be considered to interpret low-resolution data.