Semi-automated high content analysis of pollen performance using TubeTracker

Abstract

Pollen function is critical for successful plant reproduction and crop productivity and it is important to develop accessible methods to quantitatively analyze pollen performance to enhance reproductive resilience. Here we introduce TubeTracker as a method to quantify key parameters of pollen performance such as, time to pollen grain germination, pollen tube tip velocity and pollen tube survival. TubeTracker integrates manual and automatic image processing routines and the graphical user interface allows the user to interact with the software to make manual corrections of automated steps. TubeTracker does not depend on training data sets required to implement machine learning approaches and thus can be immediately implemented using readily available imaging systems. Furthermore, TubeTracker is an excellent tool to produce the pollen performance data sets necessary to take advantage of emerging AI-based methods to fully automate analysis. We tested TubeTracker and found it to be accurate in measuring pollen tube germination and pollen tube tip elongation across multiple cultivars of tomato. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=114 SRC="FIGDIR/small/624782v2_ufig1.gif" ALT="Figure 1"> View larger version (33K): org.highwire.dtl.DTLVardef@1fc2a63org.highwire.dtl.DTLVardef@42f3a2org.highwire.dtl.DTLVardef@18911d6org.highwire.dtl.DTLVardef@1f236f0_HPS_FORMAT_FIGEXP M_FIG Graphical Abstract Graphical user interface of TubeTracker showing all supported functionalities. C_FIG

Plant phenotyping relevance

植物の花粉管画像から発芽時間、伸長速度、生存性などの表現型を抽出するソフトウェア手法を開発し、複数トマト品種で精度検証しているため。

Code and data availability

The paper's authors publicly released TubeTracker, the Python software used to perform all automated pollen germination, elongation, and survival phenotyping measurements in this study, on GitHub with explicit availability language and a video sample for training.