Search

Machine learning (ML) is increasingly used in cognitive, computational and clinical neuroscience. The reliable and efficient application of ML requires a sound understanding of its subtleties and limitations. Training ML models on datasets with imbalanced classes is a particularly common problem, and it can have severe consequences if not adequately addressed. With the neuroscience ML user in mind, this paper provides a didactic assessment of the class imbalance problem and illustrates its impact through systematic manipulation of data imbalance ratios in (i) simulated data and (ii) brain data recorded with electroencephalography (EEG) and magnetoencephalography (MEG). Our results illustrate how the widely-used Accuracy (Acc) metric, which measures the overall proportion of successful predictions, yields misleadingly high performances, as class imbalance increases. Because Acc weights the per-class ratios of correct predictions proportionally to class size, it largely disregards the performance on the minority class. A binary classification model that learns to systematically vote for the majority class will yield an artificially high decoding accuracy that directly reflects the imbalance between the two classes, rather than any genuine generalizable ability to discriminate between them. We show that other evaluation metrics such as the Area Under the Curve (AUC) of the Receiver Operating Characteristic (ROC), and the less common Balanced Accuracy (BAcc) metric – defined as the arithmetic mean between sensitivity and specificity, provide more reliable performance evaluations for imbalanced data. Our findings also highlight the robustness of Random Forest (RF), and the benefits of using stratified cross-validation and hyperprameter optimization to tackle data imbalance. Critically, for neuroscience ML applications that seek to minimize overall classification error, we recommend the routine use of BAcc, which in the specific case of balanced data is equivalent to using standard Acc, and readily extends to multi-class settings. Importantly, we present a list of recommendations for dealing with imbalanced data, as well as open-source code to allow the neuroscience community to replicate and extend our observations and explore alternative approaches to coping with imbalanced data.

Co-infection with low pathogenic avian influenza virus (LPAIV) H9N2 and Newcastle disease virus (NDV) has become a worrying concern for the poultry industry. The problem arises when the hidden virus influences the replication of another suspected virus. Subsequently, misdiagnosis of the actual cause may be ended up as a source of contamination for the other healthy flocks by the spread of the covered-up virus. In this preliminary study, we determined the potential impact of concurrent infection with H9N2 and avirulent NDV (Lasota) on the virus replication in Specific Pathogen-Free embryonated chicken egg (SPF-ECE) model. Assessment of the potential interference phenomena was carried out based on embryonic lesions, mortalities, and virus replication using real-time PCR. Our results showed that H9N2 interferes with LaSota growth, regardless of which infection occurred first. Our obtained preliminary results are a call for scientists to study the interference between LPAIV H9N2 and NDV both in-vivo and in-vitro in more detail.

International audience; Avian influenza viruses of the H9N2 subtype continue to spread in wild birds and poultry worldwide. Infection with H9N2 avian influenza virus was detected for the first time in Morocco in January 2016. In this study, a total of 105 organ and tracheal swab samples from 21 broiler farms in Morocco were collected from July 2016 to October 2018 for H9N2 screening. The suspicion of disease was based on severe respiratory signs such as sneezing, coughing, rales and gasping, while H9N2 virus infection was confirmed by real-time RT-PCR. Hemagglutinin (HA) genes of four isolates were amplified by conventional RT-PCR, sequenced, and aligned for phylogenetic analyses. Among the 21 flocks, 48% (10/21) were qRT-PCR positive for H9, with the cycle threshold values ranging from18.6 to 34.8. The maximum similarity in nucleotide and protein sequences (96-98%) was observed between the Moroccan viruses and an H9 virus isolated from broiler chickens in 2017 in Burkina Faso (A/chicken/BurkinaFaso/17RS93-19/2017) and from a layer chicken in the United Arab Emirates in 2015 (A/chicken/Dubai/D2506/2015). The HA genes revealed the close relationship between the four Moroccan viruses, with 97.9%-99.9% nucleotide identity. Phylogenetic analysis showed that the Moroccan viruses belonged to the G1 lineage, and likely originated from the Middle East, as previously reported in 2016.; Virusi ptičje gripe H9N2 nastavljaju se širiti u peradi i divljih ptica širom svijeta. Infekcija niskopatogenim virusom influence H9N2 prvi je put otkrivena u Maroku u siječnju 2016. godine. U ovom je istraživanju za probir na H9N2 prikupljeno ukupno 105 organa i obrisaka iz dušnika s 21 farme brojlera od srpnja 2016. do listopada 2018. iz različitih regija Maroka. Sumnja na bolest temeljila se na teškim respiracijskim znakovima kao što su kihanje, kašljanje, hropanje i hripanje, a infekcija virusom H9N2 potvrđena je PCR-om obrnute transkripcije u stvarnom vremenu. Sekvencije gena za hemaglutinin (HA) od četiri izolata amplificirane su pomoću RT-PCR qRT-PCR poravnane za filogenetsku i analizu sličnosti aminokiselina. Od 21 uzorka jata 48 % (10/21) bilo je pozitivno na H9 s pragom broja ciklusa u rasponu od 18,6 do 34,8. Maksimalna sličnost u nukleotidnim i proteinskim sekvencijama (96 -98 %) uočena je između marokanskih virusa i virusa H9 izoliranih iz brojlerskih pilića u 2017. u Burkini Faso (A/piletina/BurkinaFaso/17RS93-19) i od kokošjeg pileta u Ujedinjenim Arapskim Emiratima u 2015. (A/piletina/Dubai/D2506/2015). HA geni otkrili su blisku vezu između četiriju virusa, s 97,9 % -99,9 % nukleotidnog identiteta. Filogenetska analiza pokazala je da marokanski virusi pripadaju lozi G1 i vjerojatno potječu s Bliskog istoka, kao što je objavljeno 2016. godine.

Avian influenza viruses of the H9N2 subtype continue to spread in wild birds and poultry worldwide. Infection with H9N2 avian influenza virus was detected for the first time in Morocco in January 2016. In this study, a total of 105 organ and tracheal swab samples from 21 broiler farms in Morocco were collected from July 2016 to October 2018 for H9N2 screening. The suspicion of disease was based on severe respiratory signs such as sneezing, coughing, rales and gasping, while H9N2 virus infection was confirmed by real-time RT-PCR. Hemagglutinin (HA) genes of four isolates were amplified by conventional RT-PCR, sequenced, and aligned for phylogenetic analyses. Among the 21 flocks, 48% (10/21) were qRT-PCR positive for H9, with the cycle threshold values ranging from18.6 to 34.8. The maximum similarity in nucleotide and protein sequences (96-98%) was observed between the Moroccan viruses and an H9 virus isolated from broiler chickens in 2017 in Burkina Faso (A/chicken/BurkinaFaso/17RS93-19/2017) and from a layer chicken in the United Arab Emirates in 2015 (A/chicken/Dubai/D2506/2015). The HA genes revealed the close relationship between the four Moroccan viruses, with 97.9%-99.9% nucleotide identity. Phylogenetic analysis showed that the Moroccan viruses belonged to the G1 lineage, and likely originated from the Middle East, as previously reported in 2016., Virusi ptičje gripe H9N2 nastavljaju se širiti u peradi i divljih ptica širom svijeta. Infekcija niskopatogenim virusom influence H9N2 prvi je put otkrivena u Maroku u siječnju 2016. godine. U ovom je istraživanju za probir na H9N2 prikupljeno ukupno 105 organa i obrisaka iz dušnika s 21 farme brojlera od srpnja 2016. do listopada 2018. iz različitih regija Maroka. Sumnja na bolest temeljila se na teškim respiracijskim znakovima kao što su kihanje, kašljanje, hropanje i hripanje, a infekcija virusom H9N2 potvrđena je PCR-om obrnute transkripcije u stvarnom vremenu. Sekvencije gena za hemaglutinin (HA) od četiri izolata amplificirane su pomoću RT-PCR qRT-PCR poravnane za filogenetsku i analizu sličnosti aminokiselina. Od 21 uzorka jata 48 % (10/21) bilo je pozitivno na H9 s pragom broja ciklusa u rasponu od 18,6 do 34,8. Maksimalna sličnost u nukleotidnim i proteinskim sekvencijama (96 -98 %) uočena je između marokanskih virusa i virusa H9 izoliranih iz brojlerskih pilića u 2017. u Burkini Faso (A/piletina/BurkinaFaso/17RS93-19) i od kokošjeg pileta u Ujedinjenim Arapskim Emiratima u 2015. (A/piletina/ Dubai/D2506/2015). HA geni otkrili su blisku vezu između četiriju virusa, s 97,9 % -99,9 % nukleotidnog identiteta. Filogenetska analiza pokazala je da marokanski virusi pripadaju lozi G1 i vjerojatno potječu s Bliskog istoka, kao što je objavljeno 2016. godine.

Newcastle disease (ND) is still a major poultry disease worldwide. Vaccination remains the principal method of controlling ND in endemic countries. Various vaccination strategies, including the use of recently developed recombinant vaccines, have been used to control it. Recombinant vaccines that use the herpesvirus of turkey (HVT) as a vector to express one of the key antigens of Newcastle disease virus (NDV) have been developed to overcome some of the drawbacks related to the use of conventional vaccines. HVT as a vector appears to have unique beneficial characteristics: it is extremely safe, it is not affected by the presence of maternally derived antibodies, and therefore can be applied in the hatchery either in ovo or to day-old chicks. Due to its persistence in the bird, the HVT vector can be expected to induce life-long immune stimulation. In the present study, the efficacy of an HVT-based vector vaccine expressing the F gene of NDV (rHVT-F) was tested against a velogenic genotype IV NDV challenge in commercial turkeys with high levels of maternal antibodies (8.7 ± 0.8 log