PaCeQuant: A Tool for High-Throughput Quantification of Pavement Cell Shape Characteristics

Microtubule-associated protein IQ67 DOMAIN5 regulates interdigitation of leaf pavement cells inArabidopsis thaliana

“…S4-S8), which is consistent with earlier reports (Möller et al, 2017; 329 Zhang et al, 2011). We applied the size thresholds of t s = 1,400 µm 2 and t m of 4,040 µm 2 , which 330 we experimentally determined in our previous work (Möller et al, 2017), to examine shape and 331 geometries in cell populations of similar sizes, referred to as small, medium and large, at 3, 5, 7 332 and 10 DAG. To distinguish between very small (tiny) and small cell populations in cotyledons 333 at 2 and 3 DAG, we included an additional size threshold of t tiny = 240 µm 2 .…”

“…PI was excited with a 555 nm laser, and emission 184 was detected between 560 and 620 nm. Segmentation, feature quantification, and graphical 185 visualization of PC shapes were conducted with the ImageJ plugin PaCeQuant and the associated 186 R script (Möller et al, 2017). For cells in cotyledons 5 to 10 DAG and in the 3rd true leaf the threshold for size filtering implemented in PaCeQuant was set to the default value of 240 µm 2 .…”

“…3A). Quantification of PC shape features 300 with PaCeQuant (Möller et al, 2017), an ImageJ-based open source tool that we developed for 301 fully automatic quantification and graphical visualization of PC shape features, revealed that cell 302 expansion was largely unaffected in iqd5 mutants, as evidenced by similar sizes of individual 303 cells and similar size distributions (Fig. 3B).…”

Mechanical stress initiates and sustains the morphogenesis of wavy leaf epidermal cells

“…Last but not least, our proposed model that is based on buckling as a mechanical event preceding cytoskeletal and cell wall polarization also provides a conceivable explanation to a phenomenon that is subtle and hence typically neglected: the presence of residual undulations in in specimens subjected to pharmacological treatments or mutations. Interference with the regulators of microtubule functioning such as brassinosteroids, regulators of ROPs such as RhoGDIs, mutations in microtubule-associated or microtubule-severing proteins such as IQD, CLASP or KATANIN all considerably decrease but in no case do they completely eliminate cell border undulations (Akita et al, 2015;Ambrose et al, 2011;Fu et al, 2005;Han et al, 2015;Kirik et al, 2007;Kotzer and Wasteneys, 2006;Liang et al, 2018;Lin et al, 2013;Liu et al, 2018;Mitra et al, 2018;Möller et al, 2017;Wu et al, 2013). The same applies to mutations or treatments affecting the actin cytoskeleton (Le et al, 2006;Rosero et al, 2016) or cellulose synthesis or crystallinity such as in any1 (Fujita et al, 2013;Ivakov and Persson, 2013) or our CGA treatments.…”

Machine and deep learning single-cell segmentation and quantification of multi-dimensional tissue images

“…Additionally, we introduce a deep learning model for classifying cell shape. We chose an autoencoder as other methods that use defined shape descriptors (Möller et al 2017) , such as circularity and concavity may not capture the relevant characteristics that a neural network can.…”