Miguel Saggu1, Ankit R Patel2, Theodoro Koulis3. 1. Late Stage Pharmaceutical Development, Genentech Inc., South San Francisco, California, 94080, USA. saggu.miguel@gene.com. 2. Late Stage Pharmaceutical Development, Genentech Inc., South San Francisco, California, 94080, USA. 3. Nonclinical Biostatistics, Genentech Inc., South San Francisco, California, 94080, USA. koulis.theodoro@gene.com.
Abstract
PURPOSE: To evaluate a random forest model that counts silicone oil droplets and non-silicone oil particles in protein formulations with large class imbalance. METHODS: In this work, we present a novel approach for automated image analysis of flow microscopy data based on random forest classification enabling rapid analysis of large data sets. The random forest approach overcomes many of the limitations of traditional classification schemes derived from simple filters or regression models. In particular, the approach does not require a priori selection of important morphology parameters. RESULTS: We analyzed silicone oil droplets and non-silicone oil particles observed in four model systems with protein concentrations of 20, 50 and 125 mg/mL. Filters based on random forests achieve higher classification accuracies when compared to regression based filters. Additionally, we showcase a procedure that allows for accurate counting of particles ≥1 μm. CONCLUSIONS: Our method is generally applicable for classification and counting of different classes of particles as long as class morphologies are differentially expressed.
PURPOSE: To evaluate a random forest model that counts silicone oil droplets and non-silicone oil particles in protein formulations with large class imbalance. METHODS: In this work, we present a novel approach for automated image analysis of flow microscopy data based on random forest classification enabling rapid analysis of large data sets. The random forest approach overcomes many of the limitations of traditional classification schemes derived from simple filters or regression models. In particular, the approach does not require a priori selection of important morphology parameters. RESULTS: We analyzed silicone oil droplets and non-silicone oil particles observed in four model systems with protein concentrations of 20, 50 and 125 mg/mL. Filters based on random forests achieve higher classification accuracies when compared to regression based filters. Additionally, we showcase a procedure that allows for accurate counting of particles ≥1 μm. CONCLUSIONS: Our method is generally applicable for classification and counting of different classes of particles as long as class morphologies are differentially expressed.
Entities:
Keywords:
classification; counting; flow microscopy; mAb; random forest; silicone oil
Authors: Sarah Zölls; Ruedeeporn Tantipolphan; Michael Wiggenhorn; Gerhard Winter; Wim Jiskoot; Wolfgang Friess; Andrea Hawe Journal: J Pharm Sci Date: 2011-12-08 Impact factor: 3.534
Authors: Satish K Singh; Nataliya Afonina; Michel Awwad; Karoline Bechtold-Peters; Jeffrey T Blue; Danny Chou; Mary Cromwell; Hans-Juergen Krause; Hanns-Christian Mahler; Brian K Meyer; Linda Narhi; Doug P Nesta; Thomas Spitznagel Journal: J Pharm Sci Date: 2010-08 Impact factor: 3.534
Authors: Daniel Weinbuch; Sarah Zölls; Michael Wiggenhorn; Wolfgang Friess; Gerhard Winter; Wim Jiskoot; Andrea Hawe Journal: J Pharm Sci Date: 2013-04-26 Impact factor: 3.534
Authors: J A Falconer; B J Naughton; D D Dunlop; E J Roth; D C Strasser; J M Sinacore Journal: Arch Phys Med Rehabil Date: 1994-06 Impact factor: 3.966
Authors: Sarah Zölls; Manuel Gregoritza; Ruedeeporn Tantipolphan; Michael Wiggenhorn; Gerhard Winter; Wolfgang Friess; Andrea Hawe Journal: J Pharm Sci Date: 2013-03-05 Impact factor: 3.534
Authors: Nils Krause; Sebastian Kuhn; Erik Frotscher; Felix Nikels; Andrea Hawe; Patrick Garidel; Tim Menzen Journal: AAPS J Date: 2021-01-04 Impact factor: 4.009