ABSTRACT
Motivation: The use of single-cell methods is expanding at an ever-increasing rate. While there are established algorithms that address cell classification, they are limited in terms of cross platform compatibility, reliance on the availability of a reference dataset and classification interpretability. Here, we introduce Pollock, a suite of algorithms for cell type identification that is compatible with popular single-cell methods and analysis platforms, provides a set of pretrained human cancer reference models, and reports interpretability scores that identify the genes that drive cell type classifications. Results: Pollock performs comparably to existing classification methods, while offering easily deployable pretrained classification models across a wide variety of tissue and data types. Additionally, it demonstrates utility in immune pan-cancer analysis. Availability and implementation: Source code and documentation are available at https://github.com/ding-lab/pollock. Pretrained models and datasets are available for download at https://zenodo.org/record/5895221. Supplementary information: Supplementary data are available at Bioinformatics Advances online.
ABSTRACT
PURPOSE: Planned flap reconstruction, allowing aggressive resections of oral cavity squamous cell carcinoma (OCSCC), may decrease positive surgical margins. The purpose of this study was to determine if length of stay (LOS), as a proxy measure for flap reconstruction, is associated with positive margin rates in OCSCC. MATERIALS AND METHODS: Data from the National Cancer Database was retrospectively collected for patients undergoing surgery for previously untreated clinical T1-3 OCSCC. Post-operative LOS was dichotomized between ≤4 and >4 days as a proxy measure for whether patients may have received flap reconstruction. Patients with LOS >4 days represent a diverse group, but those with a LOS ≤4 days are less likely to have undergone an oral cavity flap reconstruction. RESULTS: 10,107 patients were included, of which 5290 (52%) were clinical T1 and 4852 (48%) were clinical T2-3. 771 (8%) patients had a positive surgical margin. On multivariable logistic regression analysis, LOS ≤4 days was significantly associated with a positive margin resection in patients with clinical T2-3 tumors (OR 1.68, 95%CI 1.37-2.06) compared to patients with LOS >4 days. LOS was not associated with surgical margin status in patients with clinical T1 disease (OR 0.76, 95%CI 0.55-1.06). Patients with positive margin resections demonstrated worse overall survival (cT1: OR 1.35, 95%CI 1.06-1.72; cT2-3: OR 1.52, 95%CI 1.33-1.74). CONCLUSIONS: LOS >4 days after oral cavity cancer resection was significantly associated with negative surgical margins in clinical T2-3 oral cavity cancer, suggesting the possibility that patients undergoing flap reconstruction after resection have fewer positive surgical margins.
