CamDavidsonPilon/lifelines: v0.21.0

Davidson-Pilon, Cameron; Kalderstam, Jonas; Zivich, Paul N; Kuhn, Ben; Fioré-Gartland, Andrew; Moneda, Luis; Gabriel, Joseph M.; WIlson, Daniel; Parij, Alex; Stark, Kyle; Anton, Steven R.; Besson, Lilian; Jona,; Gadgil, Harsh; Golland, Dave; Hussey, Sean; Noorbakhsh, Javad; Klintberg, Andreas; Rose, Jeff; Kaluzka, Jakub; Slavitt, Isaac; Martin, Éric; Ochoa, Eduardo; Albrecht, Dylan; dhuynh,; Zgonjanin, Denis; Chen, Daniel; Fournier, Chris; Arturo,; Rendeiro, André F.

doi:10.5281/zenodo.2638135

Cited by 5 publications

(2 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Davidson-Pilon Lifelines KaplanMeierFitter (KMF) [ 42 ] Python module was used to estimate the survival function in Equation (4) and the survival curves were plotted. The KMF module required two inputs, event E and duration T, for which the patient was observed for event E. We used the ‘vital_status’ field from TCGA as event E, in order that a value of one (1) indicates death was observed while a value of zero (0) indicates right-censoring (loss to follow-up).…”

Section: Methodsmentioning

confidence: 99%

Biomarker Discovery for Meta-Classification of Melanoma Metastatic Progression Using Transfer Learning

Miñoza

Rico

Zamora

et al. 2022

Genes

View full text Add to dashboard Cite

Melanoma is considered to be the most serious and aggressive type of skin cancer, and metastasis appears to be the most important factor in its prognosis. Herein, we developed a transfer learning-based biomarker discovery model that could aid in the diagnosis and prognosis of this disease. After applying it to the ensemble machine learning model, results revealed that the genes found were consistent with those found using other methodologies previously applied to the same TCGA (The Cancer Genome Atlas) data set. Further novel biomarkers were also found. Our ensemble model achieved an AUC of 0.9861, an accuracy of 91.05, and an F1 score of 90.60 using an independent validation data set. This study was able to identify potential genes for diagnostic classification (C7 and GRIK5) and diagnostic and prognostic biomarkers (S100A7, S100A7, KRT14, KRT17, KRT6B, KRTDAP, SERPINB4, TSHR, PVRL4, WFDC5, IL20RB) in melanoma. The results show the utility of a transfer learning approach for biomarker discovery in melanoma.

show abstract

Section: Methodsmentioning

confidence: 99%

Biomarker Discovery for Meta-Classification of Melanoma Metastatic Progression Using Transfer Learning

Miñoza

Rico

Zamora

et al. 2022

Genes

View full text Add to dashboard Cite

show abstract

“…Отношение шансов (odds ratio) взаимосвязи между предикторами и вариантами исхода рассчитывали в программной среде Python с помощью пакета sklearn [6], данный пакет позволяет построить модель логистической регрессии и подобрать коэффициенты, наилучшим образом описывающие используемые данные. Отношение рисков (hazard ratio) рассчитывали с помощью пакета lifelines [7] и модуля CoxPHFitter. В статистической программной среде R созданы модели логистической регрессии с использованием пакетов Survival, MASS, car, ggplot2 [8][9][10][11][12], проведен дисперсионный анализ, далее с помощью пакета RMS [13] построены прогностическая номограмма и ROC-кривая.…”

unclassified

Nomograms for predicting local recurrence in prostate cancer patients with a positive resection margin

2022

View full text Add to dashboard Cite

Background. Prostate cancer patients with a positive resection margin after radical prostatectomy are at risk of developing local recurrence. This risk should be carefully estimated in order to choose an optimal management strategy.Aim. To develop a nomogram to predict the risk of local recurrence in patients with a positive resection margin using the data on patients who have undergone surgery.Materials and methods. Routine pathomorphological examination of surgical specimens from 2255 patients with clinically significant local and locally advanced prostate cancer revealed 364 cases of positive resection margin. Statistical analysis allowed us to identify the most significant prognostic factors. Using selected preoperative factors and a mathematical model, we created a nomogram to predict local recurrence in patients with a positive resection margin.Results. Our nomogram had an accuracy of 93% (area under the ROC curve (AUC) 0.9392; p <0.005), sensitivity of 0.99438, and specificity of 0.94545. The most significant prognostic factors included proportion of positive biopsy specimens, Gleason score (International Society of Urological Pathology (ISUP) grade estimated at routine pathomorphological examination), and presence and length of positive resection margin.Conclusion. Our mathematical model and the nomogram based on it are highly accurate for predicting local recurrence and can therefore be used for choosing an optimal management strategy.

show abstract

Combined unsupervised-supervised machine learning for phenotyping complex diseases with its application to obstructive sleep apnea

Yeol

Kim

Lee

et al. 2021

Sci Rep

View full text Add to dashboard Cite

Unsupervised clustering models have been widely used for multimetric phenotyping of complex and heterogeneous diseases such as diabetes and obstructive sleep apnea (OSA) to more precisely characterize the disease beyond simplistic conventional diagnosis standards. However, the number of clusters and key phenotypic features have been subjectively selected, reducing the reliability of the phenotyping results. Here, to minimize such subjective decisions for highly confident phenotyping, we develop a multimetric phenotyping framework by combining supervised and unsupervised machine learning. This clusters 2277 OSA patients to six phenotypes based on their multidimensional polysomnography (PSG) data. Importantly, these new phenotypes show statistically different comorbidity development for OSA-related cardio-neuro-metabolic diseases, unlike the conventional single-metric apnea–hypopnea index-based phenotypes. Furthermore, the key features of highly comorbid phenotypes were identified through supervised learning rather than subjective choice. These results can also be used to automatically phenotype new patients and predict their comorbidity risks solely based on their PSG data. The phenotyping framework based on the combination of unsupervised and supervised machine learning methods can also be applied to other complex, heterogeneous diseases for phenotyping patients and identifying important features for high-risk phenotypes.

show abstract

CamDavidsonPilon/lifelines: v0.21.0

Cited by 5 publications

References 0 publications

Biomarker Discovery for Meta-Classification of Melanoma Metastatic Progression Using Transfer Learning

Biomarker Discovery for Meta-Classification of Melanoma Metastatic Progression Using Transfer Learning

Nomograms for predicting local recurrence in prostate cancer patients with a positive resection margin

Combined unsupervised-supervised machine learning for phenotyping complex diseases with its application to obstructive sleep apnea

Contact Info

Product

Resources

About