RESUMO
BACKGROUND: Intervening in and preventing diabetes distress requires an understanding of its causes and, in particular, from a patient's perspective. Social media data provide direct access to how patients see and understand their disease and consequently show the causes of diabetes distress. OBJECTIVE: Leveraging machine learning methods, we aim to extract both explicit and implicit cause-effect relationships in patient-reported diabetes-related tweets and provide a methodology to better understand the opinions, feelings, and observations shared within the diabetes online community from a causality perspective. METHODS: More than 30 million diabetes-related tweets in English were collected between April 2017 and January 2021. Deep learning and natural language processing methods were applied to focus on tweets with personal and emotional content. A cause-effect tweet data set was manually labeled and used to train (1) a fine-tuned BERTweet model to detect causal sentences containing a causal relation and (2) a conditional random field model with Bidirectional Encoder Representations from Transformers (BERT)-based features to extract possible cause-effect associations. Causes and effects were clustered in a semisupervised approach and visualized in an interactive cause-effect network. RESULTS: Causal sentences were detected with a recall of 68% in an imbalanced data set. A conditional random field model with BERT-based features outperformed a fine-tuned BERT model for cause-effect detection with a macro recall of 68%. This led to 96,676 sentences with cause-effect relationships. "Diabetes" was identified as the central cluster followed by "death" and "insulin." Insulin pricing-related causes were frequently associated with death. CONCLUSIONS: A novel methodology was developed to detect causal sentences and identify both explicit and implicit, single and multiword cause, and the corresponding effect, as expressed in diabetes-related tweets leveraging BERT-based architectures and visualized as cause-effect network. Extracting causal associations in real life, patient-reported outcomes in social media data provide a useful complementary source of information in diabetes research.
RESUMO
BACKGROUND: TATA Binding Protein (TBP) is required for transcription initiation by all three eukaryotic RNA polymerases. It participates in transcriptional initiation at the majority of eukaryotic gene promoters, either by direct association to the TATA box upstream of the transcription start site or by indirectly localizing to the promoter through other proteins. TBP exists in solution in a dimeric form but binds to DNA as a monomer. Here, we present the first mathematical model for auto-catalytic TBP expression and use it to study the role of dimerization in maintaining the steady state TBP level. RESULTS: We show that the autogenous regulation of TBP results in a system that is capable of exhibiting three steady states: an unstable low TBP state, one stable state corresponding to a physiological TBP concentration, and another stable steady state corresponding to unviable cells where no TBP is expressed. Our model predicts that a basal level of TBP is required to establish the transcription of the TBP gene, and hence for cell viability. It also predicts that, for the condition corresponding to a typical mammalian cell, the high-TBP state and cell viability is sensitive to variation in DNA binding strength. We use the model to explore the effect of the dimer in buffering the response to changes in TBP levels, and show that for some physiological conditions the dimer is not important in buffering against perturbations. CONCLUSIONS: Results on the necessity of a minimum basal TBP level support the in vivo observations that TBP is maternally inherited, providing the small amount of TBP required to establish its ubiquitous expression. The model shows that the system is sensitive to variations in parameters indicating that it is vulnerable to mutations in TBP. A reduction in TBP-DNA binding constant can lead the system to a regime where the unviable state is the only steady state. Contrary to the current hypotheses, we show that under some physiological conditions the dimer is not very important in restoring the system to steady state. This model demonstrates the use of mathematical modelling to investigate system behaviour and generate hypotheses governing the dynamics of such nonlinear biological systems.