搜索

1 user has reported that he/she has successfully carried out the experiment using this protocol.
Comprehensive Methods for Leaf Geometric Morphometric Analyses
叶几何形态的综合分析方法   

下载 PDF 引用 收藏 提问与回复 分享您的反馈 Cited by

本文章节

参见作者原研究论文

本实验方案简略版
Plant Physiology
Mar 2016

Abstract

Leaf morphometrics are used frequently by several disciplines, including taxonomists, systematists, developmental biologists, morphologists, agronomists, and plant breeders to name just a few. Leaf shape is highly variable and can be used for identifying species or genotypes, developmental patterning within and among individuals, assessing plant health, and measuring environmental impacts on plant phenotype. Traditional leaf morphometrics requires hand tools and access to specimens, but modern efforts to digitize botanical collections make digital morphometrics a readily accessible and scientifically rigorous option. Here we provide detailed instructions for performing some of the most informative digital geometric morphometric analyses available: generalized Procrustes analysis, elliptical Fourier analysis, and shape features. This comprehensive procedure for leaf shape analysis is comprised of six main sections: A) scanning of material, B) acquiring landmarks, C) analysis of landmark data, D) isolating leaf outlines, E) analysis of leaf outlines, and F) shape features. This protocol provides a detailed reference for applying landmark and outline analysis to leaf shape as well as describing leaf shape features, thus empowering researchers to perform high throughput phenotyping for diverse applications.

Keywords: Digital leaf morphometrics (数字叶形态测量), Landmarks (标志), Generalized Procrustes Analysis (广义普鲁克分析), Leaf outlines (叶轮廓), Elliptic Fourier Descriptors (椭圆傅立叶描述符), Shape features (形状特征), Aspect ratio (纵横比), Circularity (圆度)

Background

There are a variety of approaches to digital leaf shape morphometrics, including outline or Fourier analysis, contour signatures, landmark analysis, shape features, fractal dimensions, and texture analysis (Cope et al., 2012). Among these analyses, landmark and Fourier analysis together perform exceptionally well at distinguishing between groups among leaf shapes (McLellan and Endler, 1998; Hearn, 2009). Landmark analysis is ideal for capturing aspects of shape that are consistent among all leaves within a given dataset. The selection of landmarks should include points that are biologically homologous and adequately represent the morphology of the leaf (see more pointers in Bookstein, 1991 and Zelditch et al., 2004). If leaves in the dataset do not have evolutionarily conserved shape features, ‘pseudo-landmarks’ can instead be placed (Chitwood and Sinha, 2016); that is, landmarks can be placed at equidistant points along the leaf outline relative to homologous points that act as anchors. Landmarks can then be analyzed using Generalized Procrustes Analysis (GPA), which normalizes shape data (annotated by landmarks) at equal scale, allowing for an accurate comparison of shapes regardless of their size. Outline analysis offers a more broadly applicable phenotyping method in that Elliptical Fourier Descriptors (EFDs) are used to build shape descriptors of the leaf outline (Kuhl and Giardina, 1982; Iwata and Ukai, 2002). While sensitive to noise, EFDs are ideal for large leaf datasets that have subtle differences between shapes. Shape features are an additional, simple method of outline analysis that can include the perimeter to area ratio, aspect ratio, and circularity measurements (Cope et al., 2012). In this protocol, we focus on aspect ratio and circularity, as they detect signatures of lobing and serration. Aspect ratio is the ratio of the major axis to the minor axis of a fitted ellipse, in which case values close to ‘1’ are more circular in shape regardless of lobing. Circularity is the ratio of the leaf area to perimeter outline. This measurement is useful for discriminating leaves with lobing and serration, with low circularity values indicating significant lobing and serration. This protocol is designed such that researchers can choose between all three methods (GPA, EFD, and shape features) based on which analyses best fit their data.

Materials and Reagents

  1. Herbarium specimens and/or fresh leaf material

Equipment

  1. Computer that can run Microsoft® Windows® XP (or later) and/or Mac® OS X® 10.4 (or later)
  2. Flatbed photo scanner (Epson Expression, model: 10000XL )

Software

  1. Adobe® Photoshop® CS4 (or later)
  2. Epson® Scan Utility v3.4.9.6 (https://support.epson.com/)
  3. JavaTM (https://java.com/)
  4. ImageJ (https://imagej.nih.gov/ij/)
  5. Microsoft® Excel® 2011 (or later)
  6. SHAPE v1.3 (http://lbm.ab.a.u-tokyo.ac.jp/~iwata/shape/)
    SHAPE is built for Windows but if using a Mac, Wine and Winebottler (http://winebottler.kronenberg.org/) are required
  7. R (https://r-project.org/)
    Packages: ‘shapes,’ ‘ggplot2,’ ‘devtools,’ ‘ellipse,’ and ‘roxygen2’
  8. RStudio (https://rstudio.com/products/rstudio/)
    RStudio is an optional user interface for R

Procedure

Note: Examples of R scripts and ImageJ macros referenced throughout the protocol can be freely downloaded from GitHub (link: https://github.com/htsvoboda/LeafGeometricMorphometrics.git; Note 1).

  1. Scanning fresh leaves or herbarium specimens
    1. For fresh leaves: Place leaves, with petioles removed, flat on the scanner bed. Multiple leaves can be placed on the scanning bed at once, so long as they do not overlap. If the background of the scanner is not already white, place a white piece of paper on top of the leaves; subsequent analyses work best with a solid white background (Note 2).
    2. For herbarium specimens: carefully place the entire sheet face-down on the scanner bed. Any loose material should be placed in a fragment packet on the sheet.
    3. Scans should be saved as .jpg or .tif files and named with a unique identifier (e.g., accession number).
      Note: It is best to scan at 400 dpi (or a higher resolution) as this improves the quality of the image.

  2. Acquiring and exporting landmarks
    1. Create a spreadsheet in Excel with the following column names: ‘order,’ ‘label,’ ‘x,’ and ‘y.’ This file will be referred to as the ‘master spreadsheet.’
    2. Open ImageJ
      For the first use: Select ‘Analyze > Set Measurements…’ and check only the ‘Display label’ checkbox. ‘Redirect’ and decimal place parameters can be left at their defaults.
    3. From the ImageJ menu bar, select the ‘Point selection’ tool (Figure 1; Note 3).


      Figure 1. The ImageJ menu bar. Here, the ‘Point selection’ tool button is selected.

    4. Open the first image to be landmarked. In the menu bar, select ‘File > Open...’
    5. Using the ‘Point selection’ tool in multi-point mode, begin placing landmarks on your predetermined landmark points (see Figure 2; Notes 4 and 5).


      Figure 2. Placement of landmarks on some representative leaves. A. Seventeen landmarks placed on two fresh leaves of Vitis riparia; B. Six landmarks placed on a leaf of Passiflora campanulata from a herbarium specimen.

    6. Once all landmarks have been applied to a leaf (or to multiple leaves per image), view the landmark data. In the menu bar, select ‘Analyze > Measure.’ A new window, called ‘Results’ (Figure 3), will appear with the x, y coordinates for each landmark.


      Figure 3. Example of the ‘Results’ window produced by ImageJ. This is viewed by selecting ‘Analyze > Measure’ in the menu bar.

    7. Copy (CTRL + C) and paste (CTRL + V) these data from the ‘Results’ window directly into the ‘master spreadsheet’ (see Figure 4).


      Figure 4. Example format of the ‘master spreadsheet’

    8. Close the ‘Results’ and image windows and repeat this process with each image file until all leaves have been landmarked. ‘File > Open Next’ can be used if all of the images are in the same folder.
    9. Landmark coordinates for every leaf will be pasted as-is into the ‘master spreadsheet’. The column ‘order’ may need to be adjusted at the end to reflect a continuous numerical set.
    10. Before analyzing the landmark data, the ‘master spreadsheet’ will need to be converted from Excel format (.xlsx) to a Tab Delimited Text format (.txt).
      1. Click ‘File > Save As…’ and select the .txt option.
    11. Data within the .txt file will need to be reformatted such that the rows are single leaves and the adjacent columns represent the landmark data (x, y coordinates) in sequential order. This can be done in R using the code from GitHub (file: ‘Protocol_stepB-11.R’; R Core Team, 2017).
    12. Import the reformatted file (referred to as ‘reformatted.txt’) into Excel to verify the data was properly written.
      1. Click ‘File > Import...’ and choose ‘space’ and/or ‘tab’ as the delimiter (Note 6).
    13. Supplemental columns and information (i.e., species, genotype, sex, etc.) can be added at this point to help with downstream analyses (Figure 5).


      Figure 5. The reformatted ‘master spreadsheet’ (‘reformatted.txt’) as seen in Excel. The spreadsheet now displays each row as a leaf with its landmarks distributed across the columns. Additional information can be added to the reformatted spreadsheet that may be useful in downstream analyses (e.g., columns B-F).

    14. Check landmarks for accuracy. This can be done by plotting the landmark coordinates in R with the package ‘ggplot2’ (Figure 6; Wickham, 2009). The corresponding R script can be downloaded from GitHub (file: ‘Protocol_stepB-14.R’).


      Figure 6. Landmarks are checked for accuracy using the R package ggplot. A. Example of specimen scan; B. The companion plotted landmark coordinates to the scan (A); C. Incorrectly placed landmarks will be immediately apparent visually. Note that (B) and (C) are inverted to that of (A), as pixel coordinates map inversely in a linear regression.

    15. If there appears to be an obvious error in the position of the landmarks for any image (as in Figure 6C), redo landmark placement (step B5) for the affected leaf/file and paste the updated coordinates into the ‘master spreadsheet.’ Repeat steps B11-B12 to reformat and check landmark accuracy for the new spreadsheet.

  3. Analysis of landmarks: Generalized Procrustes Analysis (GPA)
    1. To perform GPA using the R package ‘shapes,’ (Dryden, 2017) the input file must consist of only landmark data.
      1. Import (‘File > Import...’) the ‘reformatted.txt’ file into Excel.
      2. Remove all column headers and any other column information.
      3. In the menu bar, select ‘File > Save As…’ and rename the file to distinguish that it contains only x, y coordinates (e.g., ‘reformatted_coords.txt’).
    2. The data can now be processed in R using the ‘shapes’ package.
      1. The analysis produces Procrustes principal component scores and percent variance explained, Eigenleaves, Eigenvalues, among other informative outputs. The R script includes code that will write files containing PC scores and percentages for further analysis. Example R script can be found on GitHub (file: ‘Protocol_stepC-2-a.R’).
      2. Because leaf order is preserved in the output files, any additional information (e.g., individual leaf identity, species, genotype, etc.) can be re-added to these output files by adding additional columns to the file, then pasting the additional information for each leaf from the ‘reformatted.txt’ file into the PC score file.
      3. The leaves can be visualized in morphospace, using the packages ‘ggplot2,’ ‘devtools,’ (Wickham and Chang, 2015) ‘ellipse,’ (Murdoch et al., 2007) and ‘roxygen2’ (Wickham et al., 2015). Example code can be found on GitHub (file: ‘Protocol_stepC-2-c.R’). An example ordination can be seen in Figure 7.


        Figure 7. PCA ordination resulting from a Generalized Procrustes Analysis using an example landmark dataset. In this example, the leaves were labeled and color-coded according to the species identity.

  4. Isolating leaves from scans
    1. For scans of fresh leaves:
      1. Open ImageJ.
      2. Create a macro to isolate leaf images as binary images.
      3. Next, create a second macro that will select individual leaves from a scan.
        1. With the ImageJ toolbar as the active window, navigate to the menu bar and select ‘Plugins > New > Macro.’
        2. A second window will appear titled ‘Macro.txt.’ Input the text from the script found on GitHub (file: ‘Protocol_stepD-1-c-ii.txt’). Leave this macro window open. We will refer to this macro as the ‘select’ macro.
      4. Open the first image to be analyzed.
        1. With the ImageJ toolbar as the active window, navigate to the menu bar and select ‘File > Open.’
      5. Execute the ‘multiple open’ macro by highlighting all code line and clicking either ‘⌘ + R’ (for a Mac) or ‘CTRL + R’ (for a PC) (also see Note 8).
      6. The leaves within the image should now be thresholded in black and white.
      7. Isolate single leaves to extract and save as separate files by selecting the ‘Wand (tracing)’ tool from the ImageJ menu toolbar (Figure 8).


        Figure 8. The ImageJ menu toolbar. Here, the ‘Wand (tracing)’ tool is selected.

        1. Select an individual leaf by clicking on it and confirm its outline is properly highlighted.
        2. Highlight all code lines and execute the ‘select’ macro.
        3. A prompt will appear to save a binary image of a single leaf. Name the file appropriately and save in a different folder with other binary .jpeg files generated during this process for later use.
      8. Repeat steps D1a-D1h until all leaves in the image have been isolated.
      9. When each leaf in the open file has been successfully converted to binary images, add the text “run("Open Next");” as the first line of script in the ‘multiple open’ macro.
      10. Highlight all code lines, including the new line, then execute. The addition of the ‘Open Next’ command will now open the next image file in the folder.
      11. A prompt will appear to save the changes made to the current open, binary image. Do not save changes, as this will alter the original image scan.
      12. The next scan will open. Repeat the steps in D1 until all files have been converted to binary images of one leaf per image.
    2. For scans of herbarium specimens:
      1. Open Photoshop CS4.
      2. If the ‘Tools’ menu bar (Figure 9) is not already on the main screen, manually open it by selecting ‘Window > Tools.’


        Figure 9. The ‘Tools’ menu in Photoshop. A. Quick Selection Tool; B. Eyedropper Tool; C. Brush Tool; D. Crop Tool; E. Clone Stamp Tool; F. Default Foreground/Background Color Tool.

      3. From the menu, select ‘File > Open…’ and select the first scan to be processed.
      4. Identify an appropriate leaf (i.e., one that is in good condition, flat, and does not have many [or any] structures intersecting it).
      5. If the leaf is not in an upright position, rotate it using ‘Image > Image Rotation’ until the apex of the leaf is pointing upward and the base or petiole is pointing downward.
      6. Use the ‘Crop Tool’ (Figure 9D) to isolate this leaf from most of its surroundings, but leaving a buffer around each side of the leaf.
      7. If any structures (i.e., stems, tendrils, flowers, other leaves, etc.) intersect the leaf of interest, these must be removed for further analyses.
        1. Click on the ‘Default Foreground/Background Color Tool’ button (Figure 9F), making sure that ‘white’ is indicated in the foreground box (as seen in Figure 9).
        2. Select the ‘Brush Tool’ (Figure 9C) and ensure that under the Options Menu (‘Window > Options’) the ‘Mode’ is set to ‘Normal’ and the ‘Opacity’ is set to ‘100%’.
        3. Using the ‘Brush Tool’, select an appropriate brush diameter and then paint over any structures or tissues that intersect with the leaf of interest (Figure 10).


          Figure 10. An example of areas edited out using the ‘Brush Tool.’ By using this tool in Photoshop, intersecting structures have been painted over to remove them.

      8. Click on the ‘Quick Selection Tool’ (Figure 9A). Left-click on the leaf and drag along the blade until the entire lamina, and only the lamina, is outlined.
      9. Right-click in the selected area and then click ‘Select Inverse’ from the new menu.
      10. Switch back to the ‘Brush Tool’ and increase the brush diameter. Paint over the background (with white) until only the leaf remains.
      11. Right-click the image and select ‘Select Inverse’ again.
      12. Any holes, tears, or anomalies on the leaf surface need to be filled in so as to match the color of the lamina. This can be done using either the ‘Clone Stamp Tool’ (Figure 9E) or the ‘Eyedropper Tool’ (Figure 9B).
        1. To use the ‘Clone Stamp Tool,’ hold the Alt/Option key and click an intact area on the leaf. Now use the tool to fill in any damaged spots.
        2. To use the ‘Eyedropper Tool,’ left-click on an intact area of the leaf near the damaged area to extract the color. Switch to the ‘Brush Tool’ and paint over the damaged area(s) to match the color of the rest of the leaf.
      13. In some cases, especially when using .tif files, it may be necessary to ‘flatten’ the layers of the image by clicking ‘Layers > Flatten Image’ before saving. This will merge all of the edits into a single, savable image.
      14. Save each leaf (‘File > Save as…’) with a slightly different filename, but still keeping track of the original scan that it came from (e.g., ‘MO1624745_leaf1,’ ‘MO1624745_leaf2,’ etc.).
      15. Repeat steps D2a-D2n for multiple leaves per scan, and for multiple scans.
      16. To convert isolated leaf outlines to black and white images for SHAPE analysis (step D3), follow steps D1a-D1f.
    3. The program SHAPE (Iwata and Ukai, 2002) uses binary leaf outline image files in BMP format. Leaf images can be converted in ImageJ by creating a batch macro.
      1. Open ImageJ.
      2. Create a macro to easily convert many images at once.
        1. From the ImageJ menu click ‘Process > Batch > Macro...’
        2. Select the appropriate ‘Input’ folder containing the images to be converted.
        3. Select an ‘Output’ folder to contain the new binary images.
        4. Choose ‘BMP’ from the ‘Output Format’ drop-down menu.
        5. Input the text from the script found on GitHub (file: ‘Protocol_stepD-3-b-v.txt’) into the large blank space provided.
      3. View the output folder to check that the .bmp files were properly converted (Note 9).

  5. Analysis of outlines: Elliptical Fourier Descriptors (EFDs)
    1. Use the SHAPE software to convert image outlines to chain code. We encourage users to become familiar with the SHAPE User Manual in order to better explore parameter choice.
      1. Open the ChainCoder program within SHAPE.
        1. Select ‘Files > Select Image File(s).’
        2. In the following window (Figure 11), select the folder of BMP files. The BMP files will then appear in the field ‘File(s).’
        3. Select all images and select the double arrow button to transfer the desired files into the ‘Selected File(s)’ field. Press ‘OK.’


          Figure 11. Selection of .bmp files for chain coding in the software SHAPE. Use the double arrow button to move all files to the ‘Selected File(s)’ field.

      2. Before beginning the analyses, select the ‘Config’ tab (Figure 12).
        1. Set ‘Object Color’ to ‘Dark (Black),’ and ‘Scale Included’ to ‘No.’ Leave the other fields at their defaults.


          Figure 12. The ‘Config’ tab of the ChainCoder program. This allows for configuration of parameters before beginning the chain coding process.

      3. Select the ‘Processing’ tab to begin processing photos.
        1. Select ‘Load Image’ (Note 10).
        2. Deselect the ‘Select Area’ box.
        3. Select ‘Gray Scale.’
        4. Select ‘Make Histogram.’
        5. Select ‘Binarize Image.’
        6. Check the ‘Ero Dil Filter’ and ‘Dil Ero Filter’ boxes and select to what degree to ‘erode’ and ‘dilate’ the outline; this option adds and subtracts the amount of pixels from the image to give a smoother border.
        7. Select the ‘Labeling Object’ button. A new window named ‘Chain Code Data’ will appear, allowing the user to indicate which objects, over a certain number of pixels, should be isolated for analysis.
        8. Select ‘Chain Coding.’ This will add the chain code to the user selection.
        9. Select ‘Save to File.’ For the first image, you will have to name the file. Subsequently, chain codes for further images will be saved as processed (Note 11).
      4. Repeat steps E1a-E1c for all further images. If many images need to be processed, hold down the ‘enter’ key (or put a weight on it until the image processing is finished).
    2. Convert chain code to normalized EFDs.
      1. Open the CHC2NEF program within SHAPE. This program converts the chain code created above to normalized EFDs.
      2. A new window will appear (Figure 13). Select the chain code file produced in step E1c.ix (‘CHC File Name’), then create a name for the new NEF file that will be generated in the following steps (‘NEF File Name’).


        Figure 13. The CHC2NEF window. Select the chain code input file (.chc) and name the resulting normalized elliptical Fourier descriptors file (.nef).

      3. Set the ‘Max Harmonic No.’ Higher harmonic numbers lead to better shape approximations, but usually 20 is sufficient to recapitulate leaf shape accurately.
      4. Select the Normalization Method to be ‘based on the longest radius.’ This is the way the image will initially be oriented, and this option allows better subsequent manipulation to align properly.
      5. Click ‘OK.’
      6. A new window will appear. Click ‘Start !!’
      7. Orient the image so that all images are similarly aligned. This is an arbitrary designation, but needs to be consistent among all images. 
        1. There are a number of buttons to assist in this process (see Figure 14). The image can be turned by any number of degrees in either direction, and arrow buttons can be adjusted by rotation degree. Alternatively, images can rotate by 180°.


          Figure 14. Orienting leaf image chain code to normalized elliptical Fourier descriptors. Utilize the three turning buttons at the right side of the screen to orient the leaf image appropriately. NEF code will reflect these changes (bottom of window).

        2. Once aligned, click ‘Save/Next Obj.’ and repeat until all images have been normalized.
    3. Visualize Elliptical Fourier Descriptors (EFDs) using Principal Component Analysis (PCA). These analyses can be done within the SHAPE program itself (see below; Note 12).
      1. Open the ‘PrinComp’ program within SHAPE.
      2. A new window will appear. In the menu bar, select ‘Files > Open Nef File’ and select the desired .nef file.
      3. An additional window will appear (Figure 15) with parameters to determine for the PCA.


        Figure 15. The ‘NEF File Information’ window. Desired PCA parameters can be set for the normalized Elliptical Fourier Descriptors.

        1. Number of Header Lines’ is automatically set according to the NEF file.
        2. Select an appropriate number of harmonics on which to perform the PCA (the default is 20).
        3. Select the desired coefficients to keep constant.
          1) To analyze both symmetric and asymmetric variance: select ‘a-d.’
          2) To analyze symmetric variance: select ‘a’ and ‘d.’
          3) To analyze asymmetric variance: select ‘b’ and ‘c.’
      4. Perform the PCA by clicking the ‘Principal Component Analysis’ button (see Figure 16).


        Figure 16. The PrinComp program toolbar. A. ‘Principal Component Analysis’ button performs the PCA; B. ‘Calculate Principal Component Scores’ button generates PC scores for the dataset; C. ‘Reconstruct Principle Component Contours’ button generates Eigenleaves for visualization.

      5. Verify that parameters are correct in a new window, and click ‘OK.’
      6. Once the PCA results have been generated, a new window will appear to save the results file (.pcr file) in the desired folder.
      7. A new window will appear with information from the PCA (Figure 17). There are a number of tabs with useful information about the analysis.


        Figure 17. The ‘Information of Principal Component Analysis’ window. This window will appear after the program has completed the PCA run, providing useful information about the results of the analysis in various tabs.

        1. At the bottom of the window, select the ‘Make Report’ button.
        2. A new window titled ‘Report Option Dialog’ (Figure 18) will appear to select analysis information that will be written into a report (.txt file). Confirm the ‘Eigenvalues & Eigenvalue Proportions’ box is checked, as this contains the percent variance explained by each principal component (PC).


          Figure 18. Select report information for printing. Boxes checked in this window will include relevant information in a report that will be written as a .txt file.

        3. Click ‘OK.’ A new window will appear with the text file containing a PCA report.
      8. To retrieve the PC values for further analysis, click the ‘Calculate Principal Component Scores’ button in the PrinComp toolbar (Figure 16B) to create a PC score file.
        1. A new window will appear to name and select the output file. Click ‘OK.’
      9. To visualize the ‘Eigenleaves’ and what each PC represents, click the ‘Reconstruct Principal Component Contours’ button in the PrinComp toolbar (Figure 16C).
        1. A dialogue box will appear to select the number of components to reconstruct, options being ‘Reconstruct Effective Components Only’ or ‘Select Manually...’ (Figure 19). This will be the number of components that will be visualized.


          Figure 19. The ‘Reconstruct Contours’ window. In order to visualize the Eigenleaves, select the desired number of components to be reconstructed.

      10. Save the resulting ‘PC contours’ file. New windows will appear (opens automatically in SHAPE’s PrinPrint program; Figure 20), one with a graphic showing Eigenleaves, another to select draw options.
    4. The PrinPrint program can be used to view the ‘PC contours’ file at a later time.


      Figure 20. Eigenleaves visualized in the ‘PrinPrint’ program

  6. Shape features: Aspect ratio and circularity 
    1. Open ImageJ.
    2. Navigate to the menu bar and select ‘Analyze > Set Measurements...’
    3. Check the ‘Area,’ ‘Shape Descriptors,’ and ‘Display label’ boxes.
    4. Navigate to the menu bar and select ‘Process > Batch > Macro...’
      1. A new window will appear titled ‘Batch Process.’ Select the appropriate input (i.e., binary leaf image files) and output folders (Note 13).
      2. In the macro field, input the ImageJ code found on GitHub (file: ‘Protocol_stepF-4-b.txt’).
      3. Select ‘Process.’
      4. A new window will appear titled ‘Results.’ Save this report.
    5. The resulting data can be visualized using linear regression (Figure 21).


      Figure 21. Linear regression of Aspect Ratio (AR) and Circularity (Circ.) data. In this example, multiple leaves from several genotypes of Vitis riparia (purple) and V. rupestris (green) have been measured for AR and Circ., with the resulting data visualized in this linear regression. Low AR and Circ. values capture the deeper lobing and significant serration of V. riparia leaves compared to that of V. rupestris leaves.

Data analysis

Data analysis techniques are fundamental to the purpose of this protocol and are integrated within the procedure (i.e., C, E3, F5), as they are often challenging for new users to develop de novo. However, our methods represent a sampling of available methods for digital morphometric analysis. We encourage users to explore the literature and available programs to develop a method that is best suited for their material and the particular scientific inquiry.

Notes

  1. In the scripts available from GitHub, any text preceded with ‘#’ should be considered user notes–these will not be read by the computer if the whole script is copied and pasted into the R or macro consoles.
  2. Because Generalized Procrustes Analysis (GPA) and outline analysis do not require scaling, it is not necessary to include a ruler in scans for morphometric analyses; however, we recommend this as good practice.
  3. If the point tool is in the single ‘Point’ setting, simply right-click on the button to change it to the ‘Multi-point’ setting.
  4. Landmarks must be placed in the same order on all leaves.
  5. If an error was made during placement, points can be deleted by holding the CTRL key and clicking on the point. It is also possible to move points by clicking on the imprecise point, then moving it to the desired location.
  6. It may be necessary to adjust column headers or other information that might be erroneously askew from the conversion process.
  7. If there is only one fresh leaf per scan, refer to the directions in step D1e-D1f. Repeat until all images have been converted to black and white.
  8. For the first image file of the dataset, do not use the first line of code (e.g., run(“Open Next”)).
  9. BMPs are large files, and it may be convenient to use an external hard drive or cloud storage client to store and use them from this point forward.
  10. Viewing the first image in the viewing field may require using the ‘Zoom Out’ button in the top right corner to visualize properly.
  11. This produces one file for all of the images’ chain code. If chain-coding cannot be completed in one sitting, the file can be updated to include the remaining files at a later date.
  12. Alternatively, using the R package ‘Momocs’ (Bonhomme et al., 2014) converts NEFs to objects for a variety of graphical visualizations, including PCA.
  13. The desired output will not be a folder of files, rather the measurement report that can be saved as a single file. Therefore, the output folder can be the same as the input folder. To minimize file size, select a small file format (we suggest .jpeg) in the ‘Output Format’ drop down menu.

Acknowledgments

This protocol was developed in part for the publications of Chitwood et al. (2016) and Klein et al. (2017). The authors are grateful to Dr. Dan Chitwood for his comments, guidance, and expertise in using and developing these methods. We would also like to thank Dr. Allison Miller lab undergraduate members for their comments and suggestions on the protocol, as well as three reviewers who helped improve this manuscript. A Saint Louis University Graduate Research Assistantship to LLK and an Ohio University Original Work Grant to HTS supported this work.

References

  1. Bonhomme, V., Picq, S., Gaucherel, C. and Claude, J. (2014). Momocs: Outline analysis using R. J Stat Softw 56(1): 1-24.
  2. Bookstein, F. L. (1991). Morphometric Tools for Landmark Data: Geometry and Biology. Cambridge University Press, pp: 435.
  3. Chitwood, D. H., Klein, L. L., O’Hanlon, R., Chacko, S., Greg, M., Kitchen, C., Miller, A. J. and Londo, J. P. (2016). Latent developmental and evolutionary shapes embedded within the grapevine leaf. New Phytol 210(1): 343-355.
  4. Chitwood, D. H. and Sinha, N. R. (2016). Evolutionary and environmental forces sculpting leaf development. Curr Biol 26(7): R297-306.
  5. Cope, J. S., Corney, D., Clark, J. Y., Remagnino, P. and Wilkin, P. (2012). Plant species identification using digital morphometrics: A review. Expert Syst Appl 39(8): 7562–7573.
  6. Dryden, I. L. (2017). Shapes: Statistical Shape Analysis. R package version 1.1-8.
  7. Hearn, D. J. (2009). Shape analysis for the automated identification of plants from images of leaves. Taxon 58(3): 934-954.
  8. Iwata, H. and Ukai, Y. (2002). SHAPE: a computer program package for quantitative evaluation of biological shapes based on elliptic Fourier descriptors. J Heredity 93: 384-385.
  9. Klein, L. L., Caito, M., Chapnick, C., Kitchen, C., O’Hanlon Regan, Chitwood, D. H. and Miller, A. J. (2017). Digital morphometrics of two North American grapevines (Vitis: Vitaceae) quantifies leaf variation between species, within species, and among individuals. Front Plant Sci 8: 373.
  10. Kuhl, F. P. and Giardina, C. R. (1982). Elliptic features of a closed contour. Comput Vision Graphs 18: 236-258.
  11. McLellan, T. and Endler, J. A. (1998). The relative success of some methods for measuring and describing the shape of complex objects. Sys Biol 47(2): 264-281.
  12. Murdoch, D., Chow, E. D. and Celayeta, J. F. (2007). ellipse: Functions for drawing ellipses and ellipse-like confidence regions. R package version 0.3-5.
  13. R Core Team. (2017). R: a language and environment for statistical computing. R foundation for statistical computing. Vienna.
  14. Wickham, H. (2009). ggplot2: Elegant Graphics for Data Analysis. Springer.
  15. Wickham, H. and Chang, W. (2015). devtools: Tools to Make Developing R Packages Easier. R package version 1(0): 185.
  16. Wickham, H., Danenberg, P. and Eugster, M. (2015). roxygen2: In-source documentation for R, 2015. R package version 5(1).
  17. Zelditch, M. L., Swiderski, D. L., Sheets, H. D. and Fink, W. L. (2004). Geometric Morphometrics for Biologists: A Primer. Elsevier Academic Press, pp: 437.

简介

叶子形态测量经常被几个学科使用,包括分类学家,系统学家,发育生物学家,形态学家,农艺学家和植物育种者,仅举几例。叶片形状高度可变,可用于鉴定物种或基因型,个体内和个体之间的发育模式,评估植物健康以及测量对植物表型的环境影响。传统的叶子形态测量需要手工工具和标本,但现代化的数字化植物收藏的努力使数字形态测量学成为一种易于获取和科学严谨的选择。在这里,我们提供详细的说明,用于执行一些最有信息的数字几何形态测量分析:广义Procrustes分析,椭圆傅里叶分析和形状特征。叶面形状分析的综合程序由六个主要部分组成:A)材料扫描,B)获取地标,C)地标数据分析,D)隔离叶轮廓,E)叶轮廓线分析,F)形状特征。该协议提供了一个详细的参考,用于将边界和轮廓分析应用于叶子形状以及描述叶子形状特征,从而使研究人员能够为不同应用执行高吞吐量表型分析。

背景 数字叶形态测量方法有各种方法,包括轮廓或傅里叶分析,轮廓特征,地标分析,形状特征,分形维数和纹理分析(Cope等人,2012)。在这些分析中,地标和傅立叶分析在叶片形状之间的区分中表现非常好(McLellan和Endler,1998; Hearn,2009)。地标分析非常适用于捕获给定数据集中所有叶之间一致的形状方面。地标的选择应包括生物同源性并充分代表叶片形态的点(参见Bookstein,1991和Zelditch等人,2004)中的更多指针)。如果数据集中的叶子没有进化保守的形状特征,则可以将“伪地标”放置(Chitwood和Sinha,2016);也就是说,相对于作为锚点的同源点,地标可以沿着叶轮廓放置在等距点上。然后可以使用广义Procrustes分析(GPA)来分析地标,该分析(GPA)以相等的比例对形状数据(由地标注释)进行规范化,从而允许对形状进行准确比较,而不考虑其大小。轮廓分析提供了一种更广泛适用的表型方法,其中椭圆傅立叶描述符(EFD)用于构建叶轮廓的形状描述符(Kuhl和Giardina,1982; Iwata和Ukai,2002)。虽然对噪声敏感,但是EFDs对于形状之间具有微妙差异的大型叶数据集是理想的。形状特征是一种额外的,简单的轮廓分析方法,可以包括边界与面积比,纵横比和圆形度测量(Cope等人,2012)。在这个协议中,我们关注长宽比和圆形度,因为它们检测到红色和锯齿形的签名。长宽比是拟合椭圆的长轴与短轴的比率,在这种情况下接近“1”的值与圆形不同,形状更圆形。圆度是叶面积与周长轮廓的比率。该测量对于用叶片和锯齿鉴别叶片是有用的,具有低的圆形度值,表明显着的蛇纹和锯齿状。该协议的设计使得研究人员可以根据哪些分析最适合其数据来选择所有三种方法(GPA,EFD和形状特征)。

关键字:数字叶形态测量, 标志, 广义普鲁克分析, 叶轮廓, 椭圆傅立叶描述符, 形状特征, 纵横比, 圆度

材料和试剂

  1. 植物标本和/或新鲜叶材料

设备

  1. 可以运行Microsoft ® Windows ® XP(或更高版本)和/或Mac ® OS X ® 10.4的计算机(或更新)
  2. 平板照片扫描仪(Epson Expression,型号:10000XL)

软件

  1. Adobe ® Photoshop ® CS4(或更高版本)
  2. Epson ® Scan Utility v3.4.9.6( https://support.epson.com/
  3. Java TM https://java.com/
  4. ImageJ( https://imagej.nih.gov/ij/
  5. Microsoft ® Excel ® 2011(或更高版本)
  6. SHAPE v1.3( http://lbm .ab.au-tokyo.ac.jp /〜iwata/shape/
    1. SHAPE是为Windows构建的,但如果使用Mac,Wine和Winebottler( http://winebottler。 kronenberg.org/)需要
  7. R( https://r-project.org/
    1. 软件包:'形状','ggplot2','devtools','椭圆'和'roxygen2'
  8. RStudio( https://rstudio.com/products/rstudio/
    1. RStudio是R的可选用户界面

程序

注意:整个协议引用的R脚本和ImageJ宏的示例可以从GitHub中免费下载(链接: https://github.com/htsvoboda/LeafGeometricMorphometrics.git ;注1)。

  1. 扫描新鲜叶子或标本馆标本
    1. 对于新鲜的叶子:放置叶子,取下叶柄,平放在扫描仪床上。只要不重叠,可以将多个叶片立即放置在扫描床上。如果扫描仪的背景不是白色的,请将白色的纸放在叶子的顶部;随后的分析工作最适合白色背景(注2)。
    2. 对于标本标本:将整张纸张小心地放在扫描仪床上。任何松散的材料应放在片材上的片段包中。
    3. 扫描应保存为.jpg或.tif文件,并以唯一标识符(例如,登录号)命名。
      注意:最好以400 dpi(或更高分辨率)进行扫描,因为这会提高图像的质量。

  2. 获取和出口地标
    1. 在Excel中创建一个包含以下列名称的电子表格:"order","label","x"和"y"。该文件将被称为"主电子表格"。
    2. 打开ImageJ
      首次使用:选择"分析>设置测量...",并检查"显示标签"复选框。 "重定向"和小数位参数可以保留默认值。
    3. 从ImageJ菜单栏中选择"点选择"工具(图1;注3)。


      Figur e 1. ImageJ菜单栏。此处选择了"点选择"工具按钮。

    4. 打开第一个图像被标记。在菜单栏中,选择"文件>打开...'
    5. 在多点模式下使用"点选择"工具,开始在您的预定地标点上放置地标(参见图2;注4和5)。


      图2.在一些有代表性的叶子上放置地标。 A.放置在两种新鲜叶子上的十七个地标。 B.从标本馆标本放置在西番莲的叶子上的六个地标。

    6. 一旦所有的地标被应用到叶子(或每个图像的多个树叶),查看地标数据。在菜单栏中,选择"分析">"测量"。将出现一个称为"结果"(图3)的新窗口,其中包含每个地标的x,y坐标。


      图3. ImageJ生成的"结果"窗口示例。可以通过选择"Analyze>在菜单栏中测量。

    7. 复制(CTRL + C)并将这些数据从"结果"窗口粘贴(CTRL + V)直接进入"主电子表格"(见图4)。


      图4."主电子表格"的示例格式

    8. 关闭"结果"和图像窗口,并对每个图像文件重复此过程,直到所有叶子都被标记为止。 '文件>如果所有图像都在同一个文件夹中,可以使用"打开下一个"。
    9. 每个叶子的地标坐标将按原样粘贴到"主电子表格"中。可能需要在末尾调整列"顺序"以反映连续数字集。
    10. 在分析地标数据之前,"主电子表格"将需要从Excel格式(.xlsx)转换为Tab分隔文本格式(.txt)。
      1. 点击"文件>另存为...",然后选择.txt选项。
    11. 需要重新格式化.txt文件中的数据,使得行是单个叶子,相邻的列按顺序表示地标数据(x,y坐标)。这可以在R中使用GitHub的代码(文件:"Protocol_stepB-11.R"; R Core Team,2017)完成。
    12. 将重新格式化的文件(称为"重新格式化txt")导入Excel以验证数据是否正确写入。
      1. 点击"文件>导入...",并选择"空格"和/或"选项卡"作为分隔符(注6)。
    13. 此时可以添加补充栏和信息( ie ,,种类,基因型,性别,等等)以帮助下游分析(图5)。 >

      图5. Excel中显示的重新格式化的"主电子表格"('reformatted.txt')。电子表格现在将每行显示为一个叶,其地标分布在列中。可以在重新格式化的电子表格中添加其他信息,这些电子表格可能在下游分析中有用(例如,,列B-F)。

    14. 检查地标的准确性。这可以通过在R中绘制地标坐标与包ggplot2(图6; Wickham,2009)来完成。相应的R脚本可以从GitHub下载(文件:"Protocol_stepB-14.R")。


      图6.使用R包ggplot检查地标的准确性。 A.样本扫描示例。伴侣将地标坐标绘制到扫描(A); C.不正确的地标将立即在视觉上显现。注意,(B)和(C)与(A)的反转,因为像素坐标在线性回归中反向映射。

    15. 如果任何图像(如图6C所示)的地标位置似乎有明显错误,请对受影响的叶/文件重做地标位置(步骤B5),并将更新的坐标粘贴到"主电子表格"中。重复步骤B11-B12重新格式化并检查新电子表格的地标准确性。

  3. 地标分析:广义预报分析(GPA)
    1. 要使用R包"形状"(Dryden,2017)执行GPA,输入文件必须只包含地标数据。
      1. 将'reformatted.txt'文件导入('File> Import ...')到Excel中。
      2. 删除所有列标题和任何其他列信息。
      3. 在菜单栏中,选择"文件>另存为...',并重命名文件,以区分它只包含x,y坐标(例如,'reformatted_coords.txt')。
    2. 现在可以使用"形状"包在R中处理数据。
      1. 分析产生Procrustes主成分分数,百分比方差解释,特征值,特征值等信息产出。 R脚本包括将写入包含PC分数和百分比的文件以供进一步分析的代码。示例R脚本可以在GitHub上找到(文件:'Protocol_stepC-2-a.R')。
      2. 由于叶顺序保存在输出文件中,所以可以将这些输出中的任何附加信息(例如,,单个叶子身份,物种,基因型,等等等等)重新添加到这些输出中通过在文件中添加其他列,然后将每个叶片的附加信息从"reformatted.txt"文件粘贴到PC分数文件中。
      3. 使用包ggplot2,'devtools'(Wickham和Chang,2015)'椭圆'(默多克等人,2007)和'roxygen2'( Wickham等,,2015)。示例代码可以在GitHub(文件:'Protocol_stepC-2-c.R')上找到。图7中可以看到一个例子

        图7.使用示例性地标数据集的广义Procrustes分析得到的PCA排序。在此示例中,叶子根据物种标识进行了标签和颜色编码。

  4. 从扫描分离叶子
    1. 对于新鲜叶子的扫描:
      1. 打开ImageJ。
      2. 创建一个宏,将叶子图像分离为二进制图像。
      3. 接下来,创建一个第二个宏,它将从扫描中选择单独的树叶。
        1. 使用ImageJ工具栏作为活动窗口,导航到菜单栏,然后选择"插件>新>宏。"
        2. 第二个窗口将显示为"Macro.txt"。从GitHub中的脚本输入文本(文件:"Protocol_stepD-1-c-ii.txt")。让这个宏窗口打开。我们将把这个宏称为'select'宏。
      4. 打开要分析的第一个图像。
        1. 使用ImageJ工具栏作为活动窗口,导航到菜单栏,然后选择"文件>打开。"
      5. 通过突出显示所有代码行并单击"⌘+ R"(对于Mac)或"CTRL + R"(对于PC)(也见附注8))执行"多次打开"宏。
      6. 图像中的叶子现在应该被黑色和白色阈值化。
      7. 通过从ImageJ菜单工具栏中选择"Wand(跟踪)"工具(图8),将单个叶片隔离以提取并保存为单独的文件。


        图8. ImageJ菜单工具栏此处选择了"Wand(跟踪)"工具。

        1. 点击它选择一个单独的叶子,并确认其轮廓被正确突出显示。
        2. 突出显示所有代码行并执行"选择"宏。
        3. 将出现一个提示,以保存单个叶子的二进制图像。将文件命名为相应的文件夹,并保存在其他文件夹中,并在此过程中生成的其他二进制文件.jpeg文件供以后使用。
      8. 重复步骤D1a-D1h,直到图像中的所有叶子都被隔离为止。
      9. 当打开文件中的每个叶片都已成功转换为二进制图像时,添加文本"run("Open Next");"作为"多个打开"宏中的第一行脚本。
      10. 突出显示所有代码行,包括新行,然后执行。添加"打开下一个"命令现在将打开文件夹中的下一个图像文件。
      11. 将出现提示以保存对当前打开的二进制映像所做的更改。不保存更改,因为这将改变原始图像扫描。
      12. 下一次扫描将打开。重复D1中的步骤,直到所有文件都转换为每个图像一叶的二进制图像。
    2. 对于标本馆标本的扫描:
      1. 打开Photoshop CS4。
      2. 如果"工具"菜单栏(图9)尚未在主屏幕上,请通过选择"窗口">"手动"工具。"


        图9. Photoshop中的"工具"菜单。 A.快速选择工具; B.吸管工具画笔工具D.作物工具E.克隆邮票工具F.默认前景/背景颜色工具。

      3. 从菜单中选择'文件>打开..."并选择要处理的第一个扫描。
      4. 确定一个适当的叶子(,即,一个状况良好,平坦,没有很多[或任何]结构相交)。
      5. 如果叶子不是直立的位置,则使用"Image"和"图像旋转",直到叶子的顶点向上指向,底部或叶柄向下指向。
      6. 使用"作物工具"(图9D)将叶片与大部分环境隔离开来,但在叶子的每一边留下一个缓冲区。
      7. 如果任何结构( ie 。,茎,卷须,花,其他叶,)与感兴趣的叶相交,则必须删除这些结构以进行进一步的分析。 >
        1. 单击"默认前景/背景颜色工具"按钮(图9F),确保前景框中显示"白色"(如图9所示)。
        2. 选择"画笔工具"(图9C),并确保在选项菜单('窗口>选项')下,'模式'设置为'正常','不透明度'设置为'100%'。 />
        3. 使用"画笔工具",选择适当的刷子直径,然后在与感兴趣的叶子相交的任何结构或组织上绘制(图10)。


          图10.使用"画笔工具"编辑的区域示例。通过在Photoshop中使用此工具,已绘制相交结构以将其删除。

      8. 点击"快速选择工具"(图9A)。左键单击叶片并沿刀片拖动,直到整个椎板,只有椎板被勾勒出来。
      9. 右键单击所选区域,然后从新菜单中单击"选择反向"。
      10. 切换回"刷子工具"并增加刷子直径。涂上背景(用白色),直到只剩下叶子。
      11. 右键单击图像,然后再次选择"选择反向"。
      12. 需要填充叶表面上的任何孔,眼泪或异常,以匹配层的颜色。这可以使用"克隆印章工具"(图9E)或"吸管工具"(图9B)完成。
        1. 要使用"克隆邮票工具",请按住Alt/Option键,然后单击叶子上的完整区域。现在使用该工具填写任何损坏的地点。
        2. 要使用"吸管工具",请左键单击受损区域附近叶片的完整区域以提取颜色。切换到"刷子工具",并在损坏的区域上涂抹以匹配其余部分的颜色。
      13. 在某些情况下,特别是在使用.tif文件时,可能需要通过点击"图层"来"展平"图层。保存前将"平铺图像"打开。这样就可以将所有的编辑合并成一个单一的,可以破坏的图像
      14. 保存每个叶子('File>另存为...'),文件名略有不同,但仍然保持跟踪原始扫描(例如,"MO1624745_leaf1","MO1624745_leaf2" 。)。
      15. 重复步骤D2a-D2n每次扫描多个叶片,并进行多次扫描。
      16. 要将孤立的叶轮廓线转换为黑白图像进行SHAPE分析(步骤D3),请按照步骤D1a-D1f。
    3. 程序SHAPE(Iwata和Ukai,2002)使用BMP格式的二进制叶轮廓图像文件。可以通过创建批次宏来在ImageJ中转换叶子图像。
      1. 打开ImageJ。
      2. 创建一个宏以轻松转换许多图像。
        1. 从ImageJ菜单中单击"进程>批次>宏...'
        2. 选择包含要转换的图像的适当"输入"文件夹。
        3. 选择"输出"文件夹以包含新的二进制图像。
        4. 从"输出格式"下拉菜单中选择"BMP"。
        5. 从GitHub(文件:"Protocol_stepD-3-b-v.txt")中找到的脚本中输入文本到提供的大空白处。
      3. 查看输出文件夹以检查.bmp文件是否正确转换(注9)。

  5. 轮廓分析:椭圆傅立叶描述符(EFD)
    1. 使用SHAPE软件将图像轮廓转换为链码。我们鼓励用户熟悉"SHAPE用户手册",以便更好地探索参数选择。
      1. 在SHAPE中打开ChainCoder程序。
        1. 选择"文件>选择图像文件。'
        2. 在下面的窗口中(图11),选择BMP文件夹。 BMP文件将出现在"文件"字段中。
        3. 选择所有图像,然后选择双箭头按钮将所需的文件传输到"所选文件"字段。按"确定"。


          图11.在软件SHAPE中选择链接编码的.bmp文件。使用双箭头按钮将所有文件移动到"所选文件"字段。

      2. 在开始分析之前,选择"配置"选项卡(图12)。
        1. 将"对象颜色"设置为"黑色(黑色)"和"缩放包含"为"否",将其他字段设置为默认值。


          图12. ChainCoder程序的"Config"选项卡。允许在开始链式编码过程之前配置参数。

      3. 选择"处理"选项卡开始处理照片。
        1. 选择"加载图像"(注10)
        2. 取消选择"选择区域"框。
        3. 选择"灰度"。
        4. 选择"制作直方图"。
        5. 选择"二值化图像"。
        6. 检查"Ero Dil Filter"和"Dil Ero Filter"框,选择"侵蚀"和"扩大"轮廓的程度。该选项会从图像中添加和减去像素的数量,以获得更平滑的边框。
        7. 选择"标签对象"按钮。将出现一个名为"链码数据"的新窗口,允许用户指定哪些对象在一定数量的像素上应该被隔离用于分析。
        8. 选择"链式编码",这将添加链码给用户选择。
        9. 选择"保存到文件"。对于第一张图像,您必须命名该文件。随后,更多图像的链码将被保存为已处理(注11)。
      4. 对所有其他图像重复步骤E1a-E1c。如果需要处理许多图像,请按住"输入"键(或放置一个重量,直到图像处理完成)。
    2. 将链码转换为归一化EFD。
      1. 在SHAPE中打开CHC2NEF程序。该程序将上面创建的链码转换为标准化EFD。
      2. 将出现一个新窗口(图13)。选择步骤E1c.ix('CHC文件名')中生成的链码文件,然后为以下步骤("NEF文件名")中生成的新NEF文件创建一个名称。


        图13. CHC2NEF窗口选择链码输入文件(.chc),并命名生成的归一化椭圆傅立叶描述符文件(.nef)。

      3. 设置"最大谐波数"高次谐波数导致更好的形状近似,但通常20足以准确概括出叶子形状。
      4. 选择归一化方法为"基于最长半径"。这是图像最初定向的方式,此选项允许更好的后续操作来正确对齐。
      5. 点击"确定"。
      6. 将出现一个新窗口。点击"开始!!"
      7. 定向图像,使所有图像类似对齐。这是一个任意的名称,但需要在所有图像中保持一致。 
        1. 有许多按钮可以帮助这个过程(见图14)。图像可以任意方向转动任意数量,箭头按钮可以通过旋转度进行调节。或者,图像可以旋转180°。


          图14.将叶子图像链码定向到归一化的椭圆傅里叶描述符。利用屏幕右侧的三个转动按钮适当地定向叶子图像。 NEF代码将反映这些变化(窗口底部)。

        2. 一旦对齐,点击"保存/下一个对象",然后重复,直到所有图像都被归一化。
    3. 使用主成分分析(PCA)可视化椭圆傅立叶描述符(EFD)。这些分析可以在SHAPE程序本身中完成(见下文;注12)
      1. 在SHAPE中打开'PrinComp'程序。
      2. 将出现一个新窗口。在菜单栏中选择'文件>打开Nef文件"并选择所需的.nef文件。
      3. 将出现一个附加窗口(图15),其中包含用于确定PCA的参数。


        图15."NEF文件信息"窗口可以为归一化的椭圆傅里叶描述符设置所需的PCA参数。

        1. 标题行数根据NEF文件自动设置。
        2. 选择适当数量的谐波进行PCA(默认为20)。
        3. 选择所需系数保持不变。
          1)分析对称和非对称方差:选择'a-d'。
          2)分析对称方差:选择'a'和'd'。
          3)分析不对称方差:选择'b'和'c'。
      4. 通过单击"主成分分析"按钮执行PCA(参见图16)。


        图16. PrinComp程序工具栏。 A."主成分分析"按钮执行PCA; B.计算主成分分数按钮生成数据集的PC分数; C."重建原理组件轮廓"按钮可生成用于可视化的本征线。

      5. 在新窗口中验证参数是否正确,然后单击"确定"。
      6. 一旦生成PCA结果,将出现一个新窗口,将结果文件(.pcr文件)保存在所需的文件夹中。
      7. 将出现一个新窗口,其中包含PCA的信息(图17)。有许多选项卡提供有关分析的有用信息。


        图17."主成分分析信息"窗口在程序完成PCA运行后,将显示此窗口,提供有关各种选项卡中分析结果的有用信息。

        1. 在窗口底部,选择"制作报告"按钮。
        2. 将出现一个名为"报告选项对话框"(图18)的新窗口,用于选择将被写入报告(.txt文件)的分析信息。确认"特征值&特征值比例"框被检查,因为它包含由每个主要组件(PC)解释的百分比方差。


          图18.选择要打印的报告信息。此窗口中检查的框将包含将以.txt文件编写的报告中的相关信息。

        3. 单击"确定"。将出现一个新窗口,其中包含PCA报告的文本文件。
      8. 要检索PC值进行进一步分析,请单击PrinComp工具栏中的"计算主成分分数"按钮(图16B)以创建PC分数文件。
        1. 将出现一个新窗口来命名并选择输出文件。点击"确定"。
      9. 为了可视化"本征线"和每个PC表示的内容,请单击PrinComp工具栏中的"重建主体组件轮廓"按钮(图16C)。
        1. 将出现一个对话框,以选择要重构的组件数量,选项为"仅重建有效组件"或"手动选择..."(图19)。这将是可视化的组件数量。


          图19."重建轮廓"窗口为了可视化本征线,请选择要重建的所需数量的组件。

      10. 保存生成的"PC轮廓"文件。将出现新的窗口(在SHAPE的PrinPrint程序中自动打开;图20),其中一个显示Eigenleaves的图形,另一个用于选择绘图选项。
    4. PrinPrint程序可用于稍后查看"PC轮廓"文件。


      图20.在"PrinPrint"程序中可视化的本征构图

  6. 形状特征:长宽比和圆形度
    1. 打开ImageJ。
    2. 导航到菜单栏,然后选择"分析>"。设置测量...'
    3. 检查"区域","形状描述符"和"显示标签"框。
    4. 导航到菜单栏,然后选择"处理>批次>宏..."
      1. 将出现一个名为"批处理"的新窗口,选择适当的输入(,二进制叶子图像文件)和输出文件夹(注13)。
      2. 在宏字段中,输入GitHub上的ImageJ代码(文件:"Protocol_stepF-4-b.txt")。
      3. 选择'进程'。
      4. 将出现一个名为"结果"的新窗口。保存此报告。
    5. 所得到的数据可以使用线性回归进行可视化(图21)

      图21.纵横比(AR)和圆形(圆形)数据的线性回归。在本例中,来自Vitis riparia的许多基因型的多个叶子(紫色)和< EM>诉已经对AR和Circ。测量了rupestris (绿色),结果数据在此线性回归中可视化。低AR和Circ。值可以捕捉到更深的波动和显着的锯齿。与...相比。 rupestris 叶。

数据分析

数据分析技术对于本协议的目的是至关重要的,并且集成在程序( ie ,C,E3,F5)中,因为它们对于新用户往往具有挑战性,以开发从头开始。然而,我们的方法代表了用于数字形态测定分析的可用方法的抽样。我们鼓励用户探索文学和可用的方案,以开发最适合其材料和特定科学查询的方法。

笔记

  1. 在GitHub可用的脚本中,任何以'#'开头的文本都应该被视为用户注释 - 如果将整个脚本复制并粘贴到R或宏控制台中,则这些文本不会被计算机读取。
  2. 由于广义概率分析(GPA)和轮廓分析不需要缩放,因此不需要在扫描中包含尺度来进行形态测量分析;不过,我们推荐这个做法是好的做法
  3. 如果点工具位于单个"点"设置中,只需右键单击该按钮将其更改为"多点"设置。
  4. 必须在所有树叶上以相同的顺序放置地标。
  5. 如果在放置期间发生错误,可以通过按住CTRL键并点击点来删除点数。也可以通过点击不精确的点移动点,然后将其移动到所需的位置。
  6. 可能需要调整列标题或其他可能会从转换过程中歪斜的信息。
  7. 如果每次扫描只有一个鲜叶,请参阅步骤D1e-D1f中的说明。重复一遍,直到所有图像都被转换为黑白。
  8. 对于数据集的第一个图像文件,请不要使用第一行代码(例如。运行("打开下一步"))。
  9. BMP是大文件,使用外部硬盘驱动器或云端存储客户端可以方便地存储和使用它们。
  10. 在查看区域中查看第一个图像可能需要使用右上角的"缩小"按钮可以正常显示。
  11. 这将为所有图像的链码产生一个文件。如果连锁编码无法在一次会议中完成,则可以更新该文件以在以后包含剩余的文件。
  12. 或者,使用R包"Momocs"(Bonhomme等人,2014年)将NEF转换为对象,用于各种图形可视化,包括PCA。
  13. 所需的输出不会是文件夹,而是可以保存为单个文件的测量报告。因此,输出文件夹可以与输入文件夹相同。为了最小化文件大小,请在"输出格式"下拉菜单中选择一个小文件格式(我们建议.jpeg)。

致谢

该协议部分是针对Chitwood等人的出版物开发的。 (2016)和Klein等人。 (2017年)。作者非常感谢Dan Chitwood博士对使用和开发这些方法的意见,指导和专业知识。我们还要感谢艾利森·米勒博士对本科成员的实验,对协议的意见和建议,以及三位有助于改进手稿的审稿人。圣路易斯大学研究生助理LLK和俄亥俄大学HTS的原始工作授权支持这项工作。

参考

  1. Bonhomme,V.,Picq,S.,Gaucherel,C.和Claude,J。(2014)。 Momocs:使用R.的概要分析。统计软件 56(1):1-24。
  2. Bookstein,FL(1991)。< a class ="ke-insertfile"href ="http://www.cambridge.org/tw/academic/subjects/statistics-probability/statistics-life-sciences-medicine-and - 维护/形态测量 - 工具 - 地标 - 数据几何和生物学?格式= PB&isbn = 9780521585989"target ="_ blank">地标数据的形态测量工具:几何和生物学剑桥大学出版社,pp:435.
  3. Chitwood,DH,Klein,LL,O'Hanlon,R.,Chacko,S.,Greg,M.,Kitchen,C.,Miller,AJ和Londo,JP(2016)。< a class = insertfile"href ="http://www.ncbi.nlm.nih.gov/pubmed/26580864"target ="_ blank">嵌入葡萄叶中的潜在发育和进化形态 新的Phytol 210(1):343-355。
  4. Chitwood,DH和Sinha,NR(2016)。  进化和环境力量塑造叶子发育。 Curr Biol 26(7):R297-306。
  5. Cope,J.S.,Corney,D.,Clark,J.Y.,Remagnino,P.and Wilkin,P。(2012)。 使用数字形态测量的植物物种识别:评论。专家系统应用程序 39(8):7562-7573。
  6. Dryden,IL(2017)。形状:统计形状分析 R包版本 1.1-8。
  7. Hearn,DJ(2009)。< a class ="ke-insertfile"href ="http://xueshu.baidu.com/s?wd=paperuri%3A%28d6a423c860f0ee5c5d43cf78dca388fc%29&filter=sc_long_sign&tn=SE_xueshusource_2kduw22v&sc_vurl=http%3A %2F%2Fwww.jstor.org%2Fstable%2F27756958&ie = utf-8&sc_us = 7679779709164369966"target ="_ blank">用于从叶片图像自动识别植物的形状分析。 58(3):934-954。
  8. Iwata,H.和Ukai,Y.(2002)。 SHAPE:用于基于椭圆傅立叶描述符的生物形状的定量评估的计算机程序包 93:384-385。
  9. Klein,LL,Caito,M.,Chapnick,C.,Kitchen,C.,O'Hanlon Regan,Chitwood,DH和Miller,AJ(2017)。< a class ="ke-insertfile"href ="http ://journal.frontiersin.org/article/10.3389/fpls.2017.00373/full"target ="_ blank">两种北美葡萄种植的数字形态测定(Vitis :Vitaceae)量化物种间的叶片变异,种内,个人之间。 前植物科学 8:373.
  10. Kuhl,FP和Giardina,CR(1982)。  椭圆闭合轮廓的特征。计算视觉图 18:236-258。
  11. McLellan,T。和Endler,JA(1998)。测量和描述复杂物体形状的一些方法的相对成功。 Sys Biol 47(2):264-281。
  12. Murdoch,D.,Chow,ED和Celayeta,JF(2007)。  椭圆:用于绘制椭圆和类似椭圆的置信区域的函数 0.3包。
  13. R核心团队。 (2017)。 R:统计计算的语言和环境。 R基础统计计算。维也纳。
  14. Wickham,H。(2009)。< a class ="ke-insertfile"href ="http://xueshu.baidu.com/s?wd=paperuri%3A%2821a55ebe4c435834266f325e1ae41956%29&filter=sc_long_sign&tn=SE_xueshusource_2kduw22v&sc_vurl=http% 3A%2F%2Fdl.acm.org%2Fcitation.cfm%3Fid%3D2967391&ie = utf-8&sc_us = 16140654207688675768"target ="_ blank"> ggplot2:数据分析的优雅图形。 Springer
  15. Wickham,H.和Chang,W.(2015)。 devtools:使开发R包更容易的工具 R包版本 1(0):185.
  16. Wickham,H.,Danenberg,P.和Eugster,M。(2015)。< a class ="ke-insertfile"href ="https://cran.r-project.org/web/packages/roxygen2/index.html"target ="_ blank"> roxygen2:R,2015年的源代码文档 R包版本 5(1)。
  17. Zelditch,ML,Swiderski,DL,Sheets,HD and Fink,WL(2004)。  生物学家的几何形态学:一个入门。 Elsevier Academic Press ,pp:437.
  • English
  • 中文翻译
免责声明 × 为了向广大用户提供经翻译的内容,www.bio-protocol.org 采用人工翻译与计算机翻译结合的技术翻译了本文章。基于计算机的翻译质量再高,也不及 100% 的人工翻译的质量。为此,我们始终建议用户参考原始英文版本。 Bio-protocol., LLC对翻译版本的准确性不承担任何责任。
Copyright: © 2017 The Authors; exclusive licensee Bio-protocol LLC.
引用: Readers should cite both the Bio-protocol article and the original research article where this protocol was used:
  1. Klein, L. L. and Svoboda, H. T. (2017). Comprehensive Methods for Leaf Geometric Morphometric Analyses. Bio-protocol 7(9): e2269. DOI: 10.21769/BioProtoc.2269.
  2. Chitwood, D. H., Rundell, S. M., Li, D. Y., Woodford, Q. L., Yu, T. T., Lopez, J. R., Greenblatt, D., Kang, J. and Londo, J. P. (2016). Climate and Developmental Plasticity: Interannual Variability in Grapevine Leaf Morphology. Plant Physiol 170(3): 1480-1491.
提问与回复

(提问前,请先登录)bio-protocol作为媒介平台,会将您的问题转发给作者,并将作者的回复发送至您的邮箱(在bio-protocol注册时所用的邮箱)。为了作者与用户间沟通流畅(作者能准确理解您所遇到的问题并给与正确的建议),我们鼓励用户用图片或者视频的形式来说明遇到的问题。由于本平台用Youtube储存、播放视频,作者需要谷歌账户来上传视频。

当遇到任务问题时,强烈推荐您提交相关数据(如截屏或视频)。由于Bio-protocol使用Youtube存储、播放视频,如需上传视频,您可能需要一个谷歌账号。