Mining Spatio-Temporal Gene- Expression Data. Richard Baldock. MRC Human Genetics Unit, Crewe Road, Edinburgh.

Transcription

1 Mining Spatio-Temporal Gene- Expression Data Richard Baldock MRC Human Genetics Unit, Crewe Road, Edinburgh. 1

2 Development and Gene-Expression 2

3 Words are not enough mouse E12.5 nasal region Msx1 mrna 3

4 Spatio-Temporal DB for Biology QuickTime and a YUV420 codec decompressor are needed to see this picture. Spatial mapping image data gene expression database 4

5 Framework: Edinburgh Mouse Atlas (emap) QuickTime and a YUV420 codec decompressor are needed to see this picture. 5

6 6

7 7

8 8

9 emage: expression DB 9

10 emage: browse data 10

11 emage: browse entry 11

12 emage: Query by space QuickTime and a Animation decompressor are needed to see this picture. 12

13 Gene Expression Data Base Query: by both space and text... What genes are detected in spatial region at Stage X? are detected near anatomical component may be detected in may be detected near anatomical components express selected genes regions 13

14 Data Mapping Text Manual delineation Warping Radial-basis functions Piece-wise linear for efficiency Triangular/Tetrahedral mesh Manual correspondences Automatic pre-pass o Iterated Closest Point o Global-local iterated mapping 3D - manual tie-points Piece-wise mapping Constrained warping Interactive response 14

15 Data Mapping - manual 15

16 Data Mapping - Manual 16

17 Data Mapping - Manual 17

18 Data Query and Analysis Statistics & Clustering McMahon data 1356 TFs, 1800 images 11.5 dpc, TS16-19 Wholemount 168 text annotated 18

19 Data Submission EMAGE Map Submit 19

20 Data Submission EMAGE Map Submit 20

21 Data Submission EMAGE Map Submit 21

22 Summary Gene-Expression Activity 22

23 Data Query and Analysis Image Comparison Non-linear mapping and spatial pattern comparison 23

24 Data Query and Analysis Image Comparison Non-linear mapping and spatial pattern comparison 24

25 Similarity Measures Type 1 (strong): Type 3 (strong): Type 5: 25

26 Similarity Measures Type 6: 26

27 Robustness 27

28 Data Analysis - Clustering Compare each pattern with each other Use standard genomic style clustering 28

29 Data Analysis - Clustering 29

30 Data Analysis - Clustering Gray Data 30

31 Data Analysis - Clustering Gray Data 31

32 Data Analysis - Clustering Gray Data HoxC5 HoxC5 HoxD3 HoxB7 HoxB7 HoxC8 HoxC8 32

33 Data Analysis - Clustering Gray Data Oracle1/Ldb3/Cypher Meox1 myogenin (myogenic regulatory factors) eg. myogenin maintenance of Z-band cypher 33

34 Spatial Clusters 34

35 Spatial Clusters 35

36 Spatial Clusters 36

37 Spatial Clusters 37

38 Spatial Clusters 38

39 genex.hgu.mrc.ac.uk MRC Human Genetics Unit, Edinburgh EMAGE: Jeff Christiansen Shanmugasundaram Venkataraman Lorna Richardson Peter Stevenson EMAP: Albert Burger Bill Hill Nick Burton Yiya Yang Jiangao Rao Margaret Stark Julie Moss Liz Graham Allyson Ross Duncan Davidson Richard Baldock MRC e-science: Mehran Sharghi NIH GUDMAP: Derek Houghton Ying Cheng Xinguin Pi Keovillay Chanthavinout EU FP6: Guangjie Feng Nestor Milyaev Dave Clements Chris Tindall Division of Biomedical Sciences, University of Edinburgh Jonathan Bard Matt Kaufman Mouse Genome Informatics, Jackson Laboratory Martin Ringwald and the GXD team 39

40 End 40