COVID-19 Outbreak Prediction with Machine Learning

Several outbreak prediction models for COVID-19 are being used by officials around the world to make informed-decisions and enforce relevant control measures. Among the standard models for COVID-19 global pandemic prediction, simple epidemiological and statistical models have received more attention by authorities, and they are popular in the media. Due to a high level of uncertainty and lack of essential data, standard models have shown low accuracy for long-term prediction. Although the literature includes several attempts to address this issue, the essential generalization and robustness abilities of existing models needs to be improved. This paper presents a comparative analysis of machine learning and soft computing models to predict the COVID-19 outbreak. Among a wide range of machine learning models investigated, two models showed promising results (i.e., multi-layered perceptron, MLP, and adaptive network-based fuzzy inference system, ANFIS). Based on the results reported here, and due to the highly complex nature of the COVID-19 outbreak and variation in its behavior from nation-to-nation, this study suggests machine learning as an effective tool to model the outbreak.

Profiling cancer gene mutations in longitudinal epithelial ovarian cancer biopsies by targeted next-generation sequencing: a retrospective study.

BACKGROUND: The majority of patients with stage III-IV epithelial ovarian cancer (EOC) relapse after initially responding to platinum-based chemotherapy, and develop resistance. The genomic features involved in drug resistance are unknown. To unravel some of these features, we investigated the mutational profile of genes involved in pathways related to drug sensitivity in a cohort of matched tumors obtained at first surgery (Ft-S) and second surgery (Sd-S). PATIENTS AND METHODS: Matched biopsies (33) taken at Ft-S and Sd-S were selected from the 'Pandora' tumor tissue collection. DNA libraries for 65 genes were generated using the TruSeq Custom Amplicon kit and sequenced on MiSeq (Illumina). Data were analyzed using a high-performance cluster computing platform (Cloud4CARE project) and independently validated. RESULTS: A total of 2270 somatic mutations were identified (89.85% base substitutions 8.19% indels, and 1.92% unknown). Homologous recombination (HR) genes and TP53 were mutated in the majority of Ft-S, while ATM, ATR, TOP2A and TOP2B were mutated in the entire dataset. Only 2% of mutations were conserved between matched Ft-S and Sd-S. Mutations detected at second surgery clustered patients in two groups characterized by different mutational profiles in genes associated with HR, PI3K, miRNA biogenesis and signal transduction. CONCLUSIONS: There was a low level of concordance between Ft-S and Sd-S in terms of mutations in genes involved in key processes of tumor growth and drug resistance. This result suggests the importance of future longitudinal analyses to improve the clinical management of relapsed EOC.

[H1N1--pregnancy as risk factor]. / H1N1 - Risikofaktor Schwangerschaft.

There have been repeated reports of a particularly severe course of disease on H1N1 infection in pregnant women. Complemented by a review of the current literature, this paper gives an overview of recommendations with respect to vaccination, therapy and prophylaxis in this risk group.

Prenatal ultrasound diagnosis in 51 cases of holoprosencephaly: craniofacial anatomy, associated malformations, and genetics.

OBJECTIVE: To analyze the prenatal ultrasound findings of the craniofacial and extracephalic anatomy, the postnatal pathological findings, and the genetic anomalies in 51 cases of holoprosencephaly (HPE). MATERIALS AND METHODS: Between 1990 and 2005, a collective of 51 fetuses with tentative ultrasound diagnosis of HPE was recruited at two tertiary referral centers for prenatal ultrasound diagnostics via the Pia Fetal Database (GEMedical Systems, Webling, Germany). Cephalic as well as extracephalic anomalies were investigated, and all cases were subdivided into the subgroups lobar, a lobar, and semilobar HPE. In addition to the ultrasound investigation, 45 fetuses were analyzed for genetic anomalies and 21 fetuses underwent an autopsy. RESULTS: The average age at diagnosis was 21.9 weeks of gestation. There was a greater number of female fetuses, with an overall ratio of 2.67:1. In 61% of all cases, there was a reduction of growth in comparison with healthy fetuses of the same age. Within the second trimenon, the cephalic anomalies became evident when investigating the diameter of the fetal head (second trimenon: 71%below the fifth percentile; third trimenon: 92% below the fifth percentile). In 82%of the cases, extracephalic anomalies were diagnosed additionally. In 63%, the diagnosis of holoprosencephaly led to a termination of pregnancy. Ten percent of the fetuses were born alive. In 81% of the cases, the diagnosis of HPE was confirmed postnatally. The remaining 19% showed other severe cephalic and extracephalic anomalies. Chromosomal anomalies were detected in 79% of the fetuses, most frequently trisomy 13 (59%). DISCUSSION: Because of recent advances in the development and improvement of high-resolution ultrasound, early diagnosis of congenital anomalies such as HPE is now possible. In this study, which represents the largest collection of prenatally diagnosed HPE reported in the literature to date, the average age at diagnosis was earlier than in other studies. The ultrasound devices of today provide excellent images of the fetus that allow an exact diagnosis of craniomaxillofacial anomalies as well as extracephalic anomalies. Apart from a very few cases, the diagnosis of HPE is incompatible with life.

R632W mutation in PLA2G6 segregates with dystonia-parkinsonism in a consanguineous Iranian family.

BACKGROUND: PLA2G6 mutations are known to be responsible for infantile neuroaxonal dystrophy (INAD) and neurodegeneration with brain iron accumulation (NBIA). In addition, novel mutations in PLA2G6 have recently been associated with dystonia-parkinsonism in two unrelated consanguineous families. METHODS: Direct sequencing analysis of the PLA2G6 gene. RESULTS: Here, we report the segregation of R632W with disease in an Iranian consanguineous dystonia-parkinsonism pedigree. The identical mutation was previously observed in a patient affected with NBIA. CONCLUSION: We conclude that different and even identical PLA2G6 mutations may cause neurodegenerative diseases with heterogeneous clinical manifestations, including INAD, NBIA and dystonia-parkinsonism.