Data analysis of bacterial genotyping - Bionumerics Software

Transcription

1 Data analysis of bacterial genotyping - Bionumerics Software

2 Identification and Phylogeny application Bacterial, fungal and viral epidemiological typing Bacterial source tracking Mutation detection and analysis Plant and animal breeding Phylogenetic inference and evolution

3 Typing methods Traditional typing Phenotypic test panels (API, Biolog, Vitec, etc.) Serotypes Antibiotics Molecular typing 1-D electrophoresis (PFGE, RFLP, AFLP, Ribotyping, RAPD, DGGE & TGGE, etc.) Sequencing (MLST, MLVA, nucleotide or amino acid sequences, etc.) Spectrophotometer (LC-MS, SELDI)

4 Accessory machines in Bio-rad PFGE RFLP, AFLP, Ribotyping CHEF Mapper XA system Amplication and electrophoresis DGGE/TGGE Imager SELDI DCode Universal Mutation Detection System SELDI, LC-MS/MS

5 What can you get? (PFGE, RFLP, AFLP, Ribotyping, RAPD, DGGE etc.) electrophoresis

6

7 What you can store

8 What you can store

9 What can you do with Tree and Network Inference (Clustering)

10 Gel analysis I - Finding Lanes

11 Gel analysis II - Normalization

12 Gel analysis III Finding Bands

13 Selection of entries for comparison

14 Cluster analysis

15 Phylogenetic Tree analysis

16 A National laboratory dedicated to tracking foodborne infections world-wide. Each laboratory utilizes standardized genotyping methods, sharing information in real-time. The resulting surveillance provides early warning of food and waterborne disease outbreaks, emerging pathogens, and acts of bioterrorism.

17 Target organisms Campylobacter jejuni, E. coli (O157:H7), Salmonella, Shigella sonnei, Listeria monocytogenes, Vibrio cholerae, Vibrio parahemolyticus, and Yersinia pestis. Typing methods PFGE, MLVA, and SNPs (single nucleotide polymorphisms)

18 The equipments guide of the PulseNet Electrophoresis and Documentation Equipment: Bio-Rad CHEF Mapper XA System with pump and chiller module ( ) OR Bio-Rad CHEF-DR III System with pump and chiller module ( ) OR CHEF-DRII System (not desirable) Bio-Rad Gel Doc XR or good one (Chemidoc XRS, Versa 4000/5000)

19 You get a picture (PFGE, RFLP, AFLP, Ribotyping, RAPD, DGGE etc.) electrophoresis

20

21 工具列 Presence data 分析資料形式 影像檔案名 已做過的分析 菌株資料 (Entry)

22 Isolate PFGE gel Scanned tiff image Import in BioNumerics Processing image: Lane definition Normalization bands assignment Add to local database + annotations (info fields...) Analysis Comparison Matching

23 Isolate PFGE gel Scanned tiff image Import in BioNumerics Processing image: Lane definition Normalization bands assignment Add to local database + annotations (info fields...) Analysis Comparison Matching

24 Import image

25 Analysis gel

26 Analysis: Step 1 Click on the Edit Fingerprint Data icon to analyze the TIFF

27 A Fingerprint Data window should appear for your TIFF Analysis: Step 1

28 Analysis: Step 1 What if your gel looks like this? Adjust your settings in the Toolbar!!

29 Analysis: Step 1 Toolbar Overview Settings Undo Redo Add, Delete lanes Print Image Back Next Zoom In, Out Auto Search Lanes

30 Analysis: Step 1 Adjust your Settings Check the Inverted values box and click OK

31 Analysis: Step 1

32 Analysis: Step 1 Defining area of gel Move the green nodes to fit the green frame snugly around the gel lanes Hold down SHIFT and the green frame will bend Place the top of the frame just below the wells

33 Analysis: Step 1 Auto search for lanes Select the Auto Search Lanes icon Enter the number of lanes on the TIFF

34 Analysis: Step 1 Fingerprint Settings Adjust individual lane width OR Adjust width of all lanes under Settings

35 Analysis: Step 1 Click on the Raw Data tab Select the thickness for the gel strips Click OK

36 Analysis: Step 1 1) 2) 3) The frame of each gel strip should not 1) cut off the edges of the lane, nor 2) include too much space, but instead should 3) include all of the lane

37 In the Fingerprint Data window under the Edit menu select Edit Tone Curve Analysis: Step 1

38 Analysis: Step 1 On the Gel Tone Curve screen click the Linear button to optimize image Before After Before Optimization After Optimization The Enhance weak bands and Enhance dark bands buttons can be used to further optimize gel with weak or dark bands

39 Analysis: Step 1 Click the Next arrow On to Step 2!!

40 Analysis: Step 2 Defining densitometric curves Use the blue node to drag the strip to the best area of each lane, avoiding artifacts (specks, etc.) on the gel Densitometric curve Peaks correspond with band intensity in gel lanes

41 Objectives I. Copying a TIFF to the database II. Analyze the TIFF a. Converting a TIFF to gel strips b. Normalizing the gel c. Finding gel bands d. Linking database entries to gel lanes

42 Analysis: Step 3 Selecting Standard Lanes 1) Select the standard or reference lanes 2) Select the icon to designate the standard lanes Note: At minimum for proper normalization, the first, last, and every 4 th or 5 th lane should be a standard lane!!

43 Analysis: Step 3 Marking to the Global Standard 3 Ways to mark reference bands 1. Manual band assign only 2. Auto band assign only 3. Combination Auto and Manual band selection

44 Analysis: Step 3 Manual band assign only a) Click on its corresponding global standard lane band marker NOTE: Doublet in Standard Lanes are marked differently for different organisms b) Click on the corresponding band in the reference lane and press Enter

45 Analysis: Step 3 Auto band assign only a) Click the icon to auto assign reference bands NOTE: Doublet in the standard lane will not be marked b) Click Ok and standards will be automatically marked c) Check band assignments

46 Analysis: Step 3 Combination Auto and Manual band assignment a) Mark top, bottom and middle bands b) Click the Auto assign icon c) Keep existing band assignments by checking the box and clicking OK NOTE: Doublet in Standard Lanes are marked differently for different organisms

47 Analysis: Step 3 Showing Distortion Bars Click on the normalization icon to see the normalized view SAVE YOUR WORK!!! Click on Show distortion bars to check normalization

48 Analysis: Step 3 Showing Distortion Bars Light color distortion bars in the gel indicate good normalization Dark colors, especially in one part of the gel indicates poor normalization, possibly due to an incorrect band assignment

49 Objectives I. Copying a TIFF to the database II. Analyze the TIFF a. Converting a TIFF to gel strips b. Normalizing the gel c. Finding gel bands d. Linking database entries to gel lanes

50 Analysis: Step 4 2 Ways to Mark Gel Bands 1. Auto Band assign 2. Manual Band assign

51 Analysis: Step 4 Auto Assign Bands Adjust to alter band detection Preview band assignments Assign bands to all or one lane Click Yes to preserve band assignments

52 Analysis: Step 4 Manual Band Assignment Click on each individual band and mark it by pressing the ENTER key

53 Analysis: Step 4 Verify Bands Check over your work Add or delete bands where needed by selecting the band(s) and pressing the ENTER or DELETE keys Single Band Double Band Double, overlapping bands are denoted with arrows on the ends of the band

54 Analysis: Step 4 Save and Close Click on the Save icon Click Yes to save changes Click No to preserve Default settings

55

56 Analysis gel

57 Tips for analysis Use a picture of your gel to help identify bands Use the zoom (in/out) buttons for ease of viewing bands

58 Objectives I. Copying a TIFF to the database II. Analyze the TIFF a. Converting a TIFF to gel strips b. Normalizing the gel c. Finding gel bands d. Linking database entries to gel lanes

59 Linking Lanes to Database Entries: Changing Fingerprint Type Add all lane to database If the gel contains lanes restricted with multiple enzymes, the fingerprint type for those lanes should be designated before linking To change: -Right-click on the lane -Select Change fingerprint type

60 Linking Lanes to Database Entries: Changing Fingerprint Type

61 Linking Lanes to Database Entries: Creating a link 2 Ways to Link 1) Select the lane and click the icon or 2) Right-click on the lane and select Link Lane DO NOT LINK STANDARD LANES!!

62 Linking Lanes to Database Entries: Creating a new key Enter the isolate number or key and click OK Since we are creating a new entry, select Yes

63 Tips for good database Setup a SOP for Running condition Reagent Making gel Environment The guideline of analysis

64 Tips for good database 1.5 cm

65 Results (Time course) hr.

66 Isolate PFGE gel Scanned tiff image Import in BioNumerics Processing image: Lane definition Normalization bands assignment Add to local database + annotations (info fields...) Analysis Comparison Matching

67 Add new information field

68 Double click Add new information

69

70 Isolate PFGE gel Scanned tiff image Import in BioNumerics Processing image: Lane definition Normalization bands assignment Add to local database + annotations (info fields...) Analysis Comparison Matching

71 Selection of entries for comparison Standard BioNumerics Query tool Manual selection: CTRL+mouse click on entries SHIFT+mouse click on range of entries

72 Standard BioNumerics query tool Fill in the fields required for the query all fields available combinations possible ( AND )

73 Create a comparison with the current selection Current selection after query

74 Visualise data in comparison 1.Press the button of the experiment you want to analysis You could see the PFGE results by press the image button of the experiment you want to see

75 Visualise data in comparison Press the bands button to show the bands of the currently displayed experiment

76 Visualise data in comparison

77 Perform cluster analysis Press the cluster analysis button Use Calculate cluster analysis

78 Perform cluster analysis

79 Band based similarities Most common coefficients: Jaccard = N common N total Dice = 2 x N common = Nei & Li = N common + N total 2[AB] [A] + [B] Jeffrey s x = N common N A + N common N B

80 Perform cluster analysis

81 Detailed comparison of two entries - At any time, two entries can be compared Select the two entries you want to compare Press Ctrl+Alt+C to open pairwize comparison (or Comparison>Compare two entries) in main window

82 Detailed comparison of two entries (Ctrl+Alt+C) Select experiment to visualise

83 Phylogenetic Tree analysis The limitation of molecular typing methods?

84 Composite Data sets

85 Case 1 New probes used for IS1245 and IS1311 restriction fragment length polymorphism of Mycobacterium avium subsp. avium and Mycobacterium avium subsp. hominissuis isolates of human and animal origin in Norway Tone Bjordal Johansen, Ingrid Olsen, Merete Rusås Jensen, Ulf R Dahle, Gudmund Holstad and Berit Djønne BMC Microbiology 2007, 7:14

86 Background Mycobacterium avium is an environmental mycobacterium that can be divided into the subspecies avium, hominissuis, paratuberculosis and silvaticum. Some M. avium subspecies are opportunistic pathogens for animals and humans. Isolates of M. avium originating from humans (n = 37), pigs (n = 51) and wild birds (n = 10) in Norway, total 98 isolates They were examined by IS1245 and IS1311 RFLP using new and specific probes and for the presence of IS901 and ISMpa1 by PCR

87 RFLP

88 11 isolates identical in their banding pattern by IS1311 RFLP, which could be further separated by IS1245 RFLP.

89 A dendrogram showing the cluster analysis of the composite dataset of RFLP experiments Combained two probes using to track IS1311 and the isolates of different IS1245 orgainsas probes with the certain genotypings A total of 14 clusters (defined by 80% similarity) are grouped in rectangular frames and labelled A-N.

90 Isolate PFGE gel Scanned tiff image Import in BioNumerics Processing image: Lane definition Normalization bands assignment Add to local database + annotations (info fields...) Analysis Comparison Matching composite data

91 Composite Data sets

92 How to combine data meaningfully?

93 Option 1: averaging similarity matrices

94 Creating a Composite Data Set Choose Experiments Create new composite data set Or, right-click under the Experiments window Right-click

95 Creating a Composite Data Set 1.Enter the name of your composite data set and click OK 2.Highlight the experiment/character types you want included 3.Choose Use in composite data set from the Experiment window

96 Creating a Composite Data Set Choose Correct for internal weights from the Experiments window Notice new Composite Data Set in Experiments area

97 Comparisons using Composite Data Sets We bring up both sets of images, and then make sure the Composite Data Set is highlighted

98 Comparisons using Composite Data Sets 1 2 1) Choose Calculate cluster analysis under the dendrogram icon 2) Under Similarity, choose Average from experiments

99 Resulting Dendrogram

100 Conclusion Different experiment could combine in one analysis and calculate the congruence. The Bionumerics could analyze more than 10,000 quarys, even the experiments got in different time and labs. The Bionumerics help for tracking infections world-wide.

101 Final warning A computer is just a tool! Final decision concerning resemblance between fingerprints = the human eye The computer can help to Store the data Process the data (normalisation) Rapidly search massive amounts of information (queries & matching)

102 Thank you for attention!