BACKGROUND: Tumours contain hypoxic regions that select for an aggressive cell phenotype; tumour hypoxia induces metastasisassociated genes. Treatment refractory patients with metastatic cancer show increased numbers of circulating tumour cells (CTCs), which are also associated with disease progression. The aim of this study was to examine the as yet unknown relationship between hypoxia and CTCs. METHODS: We generated human MDA-MB-231 orthotopic xenografts and, using a new technology, isolated viable human CTCs from murine blood. The CTCs and parental MDA-MB-231 cells were incubated at 21 and 0.2% (hypoxia) oxygen, respectively. Colony formation was assayed and levels of hypoxia-and anoxia-inducible factors were measured. Xenografts generated from CTCs and parental cells were compared. RESULTS: MDA-MB-231 xenografts used to generate CTCs were hypoxic, expressing hypoxia factors: hypoxia-inducible factor1 alpha (HIF1a) and glucose transporter protein type 1 (GLUT1), and anoxia-induced factors: activating transcription factor 3 and 4 (ATF3 and ATF4). Parental MDA-MB-231 cells induced ATF3 in hypoxia, whereas CTCs expressed it constitutively. Asparagine synthetase (ASNS) expression was also higher in CTCs. Hypoxia induced ATF4 and the HIF1a target gene apelin in CTCs, but not in parental cells. Hypoxia induced lower levels of carbonic anhydrase IX (CAIX), GLUT1 and BCL2/adenovirus E1B 19-KD protein-interacting protein 3 (BNIP3) proteins in CTCs than in parental cells, supporting an altered hypoxia response. In chronic hypoxia, CTCs demonstrated greater colony formation than parental cells. Xenografts generated from CTCs were larger and heavier, and metastasised faster than MDA-MB-231 xenografts. CONCLUSION: CTCs show an altered hypoxia response and an enhanced aggressive phenotype in vitro and in vivo.

Pretrained embedding representations of biological sequences which capture meaningful properties can alleviate many problems associated with supervised learning in biology. We apply the principle of mutual information maximization between local and global information as a self-supervised pretraining signal for protein embeddings. To do so, we divide protein sequences into fixed size fragments, and train an autoregressive model to distinguish between subsequent fragments from the same protein and fragments from random proteins. Our model, CPCProt, achieves comparable performance to state-of-the-art self-supervised models for protein sequence embeddings on various downstream tasks, but reduces the number of parameters down to 0.9% to 8.9% of benchmarked models. Further, we explore how downstream assessment protocols affect embedding evaluation, and the effect of contrastive learning hyperparameters on empirical performance. We hope that these results will inform the development of contrastive learning methods in protein biology and other modalities.

Writing a good essay typically involves students revising an initial paper draft after receiving feedback. We present eRevise, a web-based writing and revising environment that uses natural language processing features generated for rubricbased essay scoring to trigger formative feedback messages regarding students' use of evidence in response-to-text writing. By helping students understand the criteria for using text evidence during writing, eRevise empowers students to better revise their paper drafts. In a pilot deployment of eRevise in 7 classrooms spanning grades 5 and 6, the quality of text evidence usage in writing improved after students received formative feedback then engaged in paper revision.

Clinical machine learning models experience significantly degraded performance in datasets not seen during training, e.g., new hospitals or populations. Recent developments in domain generalization offer a promising solution to this problem by creating models that learn invariances across environments. In this work, we benchmark the performance of eight domain generalization methods on multi-site clinical time series and medical imaging data. We introduce a framework to induce synthetic but realistic domain shifts and sampling bias to stress-test these methods over existing nonhealthcare benchmarks. We find that current domain generalization methods do not achieve significant gains in out-of-distribution performance over empirical risk minimization on real-world medical imaging data, in line with prior work on general imaging datasets. However, a subset of realistic induced-shift scenarios in clinical time series data exhibit limited performance gains. We characterize these scenarios in detail, and recommend best practices for domain generalization in the clinical setting. CCS CONCEPTS• Computing methodologies → Machine learning; • Applied computing → Health informatics; • General and reference → Empirical studies.

In this work, we examine the extent to which embeddings may encode marginalized populations differently, and how this may lead to a perpetuation of biases and worsened performance on clinical tasks. We pretrain deep embedding models (BERT) on medical notes from the MIMIC-III hospital dataset, and quantify potential disparities using two approaches. First, we identify dangerous latent relationships that are captured by the contextual word embeddings using a fill-in-the-blank method with text from real clinical notes and a log probability bias score quantification. Second, we evaluate performance gaps across different definitions of fairness on over 50 downstream clinical prediction tasks that include detection of acute and chronic conditions. We find that classifiers trained from BERT representations exhibit statistically significant differences in performance, often favoring the majority group with regards to gender, language, ethnicity, and insurance status. Finally, we explore shortcomings of using adversarial debiasing to obfuscate subgroup information in contextual word embeddings, and recommend best practices for such deep embedding models in clinical settings.

Psychosocial service utilization by outpatients with bipolar disorder is strongly linked to greater severity/complexity of illness. Potential moderators, such as insurance status and availability of care, should be examined in future studies.

Manually grading the Response to Text Assessment (RTA) is labor intensive. Therefore, an automatic method is being developed for scoring analytical writing when the RTA is administered in large numbers of classrooms. Our long-term goal is to also use this scoring method to provide formative feedback to students and teachers about students' writing quality. As a first step towards this goal, interpretable features for automatically scoring the evidence rubric of the RTA have been developed. In this paper, we present a simple but promising method for improving evidence scoring by employing the word embedding model. We evaluate our method on corpora of responses written by upper elementary students.

Chitosan-based activated carbon with high specific surface area (approximately 3500 m2/g) was obtained by a two-step activation process. The effects of preparation parameters such as the impregnation ratio, activation temperature and activation time on the surface area and pore structure of those chitosan-based activated carbons were studied by nitrogen adsorption at 77 K. The crystallinity degree of the samples was analyzed by X-ray diffraction and the surface morphology of char and activated carbon was observed using TEM. And the cyclic voltammetry measurement showed that the activated carbon with highest surface area exhibited a super capacitance of 338 F/g at the scan rate of 2 mV/s, and the corresponding charge-discharge curves were symmetrical triangle, indicating that the electrode material obtained at the optimum conditions has good performance in electrochemical stability and reversibility.