Inflammatory stimuli have divergent effects on peripheral platelet counts, although the mechanisms of thrombocytopenic and thrombocytotic responses remain poorly understood. A candidate gene approach targeting 326 polymorphic genes enriched in thrombopoietic and cytokine signaling pathways was applied to identify single nucleotide variants (SNVs) implicated in enhanced platelet responses in cohorts with reactive thrombocytosis (RT) or essential (myeloproliferative neoplasm [MPN]) thrombocytosis (ET). Cytokine profiles incorporating a 15-member subset, pathway topology, and functional interactive networks were distinct between ET and RT, consistent with distinct regulatory pathways of exaggerated thrombopoiesis. Genetic studies using aggregate (ET + RT) or ET-restricted cohorts identified associations with 2 IFNA16 (interferon-α16) SNVs, and the ET associations were validated in a second independent cohort (P = .0002). Odds ratio of the combined ET cohort (n = 105) was 4.92, restricted to the JAK2V617F-negative subset (odds ratio, 5.01). ET substratification analysis by variant IFNA16 exhibited a statistically significant increase in IFN-α16 levels (P = .002) among 16 quantifiable cytokines. Recombinantly expressed variant IFN-α16 encompassing 3 linked non-synonymous SNVs (E65H95P133) retained comparable antiviral and pSTAT signaling profiles as native IFN-α16 (V65D95A133) or IFN-α2, although both native and variant IFN-α16 showed stage-restricted differences (compared with IFN-α2) of IFN-regulated genes in CD34+-stimulated megakaryocytes. These data implicate IFNA16 (IFN-α16 gene product) as a putative susceptibility locus (driver) within the broader disrupted cytokine network evident in MPNs, and they provide a framework for dissecting functional interactive networks regulating stress or MPN thrombopoiesis.

Due to spectacular gains during periods of rapid price increase and unpredictably large drops, Bitcoin has become a popular emergent asset class over the past few years. In this paper, we are interested in predicting the crashes of Bitcoin market. To tackle this task, we propose a framework for deep learning time series classification based on contrastive learning. The proposed framework is evaluated against six machine learning (ML) and deep learning (DL) baseline models, and outperforms them by 15.8% in balanced accuracy. Thus, we conclude that the contrastive learning strategy significantly enhance the model’s ability of extracting informative representations, and our proposed framework performs well in predicting Bitcoin crashes.