1.1 Binary file grant.doc has changed

2.1 --- a/grant.html Mon Apr 13 19:38:30 2009 -0700 2.2 +++ b/grant.html Mon Apr 13 20:27:32 2009 -0700 2.3 @@ -8,8 +8,8 @@ 2.4 (2) develop an algorithm to suggest new ways of carving up a structure into anatomical subregions, based on spatial 2.5 patterns in gene expression 2.6 (3) create a 2-D “flat map” dataset of the mouse cerebral cortex that contains a flattened version of the Allen Mouse 2.7 -Brain Atlas ISH data, as well as the boundaries of cortical anatomical areas. Use this dataset to validate the methods 2.8 -developed in (1) and (2). 2.9 +Brain Atlas ISH data, as well as the boundaries of cortical anatomical areas. This will involve extending the functionality of 2.10 +Caret, an existing open-source scientific imaging program. Use this dataset to validate the methods developed in (1) and (2). 2.11 In addition to validating the usefulness of the algorithms, the application of these methods to cerebral cortex will produce 2.12 immediate benefits, because there are currently no known genetic markers for many cortical areas. The results of the project 2.13 will support the development of new ways to selectively target cortical areas, and it will support the development of a 2.14 @@ -243,9 +243,17 @@ 2.15 underneath each pixel, with red meaning a lot of expression and blue meaning little. 2.16 Preliminary work 2.17 Format conversion between SEV, MATLAB, NIFTI 2.18 -todo 2.19 +We have created software to (politely) download all of the SEV files from the Allen Institute website. We have also created 2.20 +software to convert between the SEV, MATLAB, and NIFTI file formats, as well as some of Caret’s formats. 2.21 Flatmap of cortex 2.22 -todo 2.23 +We created a mask which selects only those voxels within the ABA atlas space which belong to cerebral cortex. 2.24 +todo 2.25 +Using Caret, [1] 2.26 +We manually entered the boundaries of each cortical area into Caret. 2.27 +Cortical layers are found at different depths in different parts of the cortex. We have manually demarcated the depth of 2.28 +the outer boundary of cortical layer 5 throughout the cortex. 2.29 +In preparation for extracting the layer-specific datasets, we have extended Caret with routines that allow the depth of 2.30 +the ROI for volume-to-surface projection to vary. 2.31 Using combinations of multiple genes is necessary and sufficient to delineate some cortical areas 2.32 Here we give an example of a cortical area which is not marked by any single gene, but which can be identified combi- 2.33 natorially. according to logistic regression, gene wwc19 is the best fit single gene for predicting whether or not a pixel on 2.34 @@ -260,24 +268,19 @@ 2.35 Conditional entropy todo 2.36 Gradient similarity todo 2.37 Geometric and pointwise scoring methods provide complementary information 2.38 -To show that local geometry can provide useful information that cannot be detected via pointwise analyses, consider Fig. 2.39 -. The top row of Fig. displays the 3 genes which most match area AUD, according to a pointwise method11. The bottom 2.40 -row displays the 3 genes which most match AUD according to a method which considers local geometry12 The pointwise 2.41 -method in the top row identifies genes which express more strongly in AUD than outside of it; its weakness is that this 2.42 _________________________________________ 2.43 9“WW, C2 and coiled-coil domain containing 1”; EntrezGene ID 211652 2.44 10“mitochondrial translational initiation factor 2”; EntrezGene ID 76784 2.45 - 11For each gene, a logistic regression in which the response variable was whether or not a surface pixel was within area AUD, and the predictor 2.46 -variable was the value of the expression of the gene underneath that pixel. The resulting scores were used to rank the genes in terms of how well 2.47 -they predict area AUD. 2.48 - 12For each gene the gradient similarity (see section ??) between (a) a map of the expression of each gene on the cortical surface and (b) the 2.49 -shape of area AUD, was calculated, and this was used to rank the genes. 2.50 2.51 2.52 2.53 Figure 2: The top row shows the three genes which (individually) best predict area AUD, according to logistic regression. 2.54 The bottom row shows the three genes which (individually) best match area AUD, according to gradient similarity. From 2.55 left to right and top to bottom, the genes are Ssr1, Efcbp1, Aph1a, Ptk7, Aph1a again, and Lepr 2.56 +To show that local geometry can provide useful information that cannot be detected via pointwise analyses, consider Fig. 2.57 +. The top row of Fig. displays the 3 genes which most match area AUD, according to a pointwise method11. The bottom 2.58 +row displays the 3 genes which most match AUD according to a method which considers local geometry12 The pointwise 2.59 +method in the top row identifies genes which express more strongly in AUD than outside of it; its weakness is that this 2.60 includes many areas which don’t have a salient border matching the areal border. The geometric method identifies genes 2.61 whose salient expression border seems to partially line up with the border of AUD; its weakness is that this includes genes 2.62 which don’t express over the entire area. Genes which have high rankings using both pointwise and border criteria, such as 2.63 @@ -285,6 +288,8 @@ 2.64 we deliberately chose a “difficult” area in order to better contrast pointwise with geometric methods. 2.65 Areas which can be identified by single genes 2.66 todo 2.67 +Areas can sometimes be marked by underexpression 2.68 +todo 2.69 Specific to Aim 1 (and Aim 3) 2.70 Forward stepwise logistic regression todo 2.71 SVM on all genes at once 2.72 @@ -299,13 +304,18 @@ 2.73 Specific to Aim 2 (and Aim 3) 2.74 Raw dimensionality reduction results 2.75 todo 2.76 +_________________________________________ 2.77 + 11For each gene, a logistic regression in which the response variable was whether or not a surface pixel was within area AUD, and the predictor 2.78 +variable was the value of the expression of the gene underneath that pixel. The resulting scores were used to rank the genes in terms of how well 2.79 +they predict area AUD. 2.80 + 12For each gene the gradient similarity (see section ??) between (a) a map of the expression of each gene on the cortical surface and (b) the 2.81 +shape of area AUD, was calculated, and this was used to rank the genes. 2.82 + 135-fold cross-validation. 2.83 (might want to incld nnMF since mentioned above) 2.84 Dimensionality reduction plus K-means or spectral clustering 2.85 Many areas are captured by clusters of genes 2.86 todo 2.87 todo 2.88 -_________________________________________ 2.89 - 135-fold cross-validation. 2.90 Research plan 2.91 todo amongst other things: 2.92 Develop algorithms that find genetic markers for anatomical regions

3.1 Binary file grant.odt has changed

4.1 Binary file grant.pdf has changed

5.1 --- a/grant.txt Mon Apr 13 19:38:30 2009 -0700 5.2 +++ b/grant.txt Mon Apr 13 20:27:32 2009 -0700 5.3 @@ -9,7 +9,7 @@ 5.4 5.5 (2) develop an algorithm to suggest new ways of carving up a structure into anatomical subregions, based on spatial patterns in gene expression\\ 5.6 5.7 -(3) create a 2-D "flat map" dataset of the mouse cerebral cortex that contains a flattened version of the Allen Mouse Brain Atlas ISH data, as well as the boundaries of cortical anatomical areas. Use this dataset to validate the methods developed in (1) and (2).\\ 5.8 +(3) create a 2-D "flat map" dataset of the mouse cerebral cortex that contains a flattened version of the Allen Mouse Brain Atlas ISH data, as well as the boundaries of cortical anatomical areas. This will involve extending the functionality of Caret, an existing open-source scientific imaging program. Use this dataset to validate the methods developed in (1) and (2).\\ 5.9 5.10 In addition to validating the usefulness of the algorithms, the application of these methods to cerebral cortex will produce immediate benefits, because there are currently no known genetic markers for many cortical areas. The results of the project will support the development of new ways to selectively target cortical areas, and it will support the development of a method for identifying the cortical areal boundaries present in small tissue samples. 5.11 5.12 @@ -175,10 +175,26 @@ 5.13 == Preliminary work == 5.14 5.15 === Format conversion between SEV, MATLAB, NIFTI === 5.16 -todo 5.17 +We have created software to (politely) download all of the SEV files from the Allen Institute website. We have also created software to convert between the SEV, MATLAB, and NIFTI file formats, as well as some of Caret's formats. 5.18 + 5.19 5.20 === Flatmap of cortex === 5.21 -todo 5.22 +We created a mask which selects only those voxels within the ABA atlas space which belong to cerebral cortex. 5.23 + 5.24 +todo 5.25 + 5.26 +Using Caret, \cite{van_essen_integrated_2001} 5.27 + 5.28 +We manually entered the boundaries of each cortical area into Caret. 5.29 + 5.30 +Cortical layers are found at different depths in different parts of the cortex. We have manually demarcated the depth of the outer boundary of cortical layer 5 throughout the cortex. 5.31 + 5.32 +In preparation for extracting the layer-specific datasets, we have extended Caret with routines that allow the depth of the ROI for volume-to-surface projection to vary. 5.33 + 5.34 + 5.35 + 5.36 + 5.37 + 5.38 5.39 5.40 5.41 @@ -228,6 +244,9 @@ 5.42 5.43 todo 5.44 5.45 +\vspace{0.3cm}**Areas can sometimes be marked by underexpression** 5.46 + 5.47 +todo 5.48 5.49 === Specific to Aim 1 (and Aim 3) === 5.50 \vspace{0.3cm}**Forward stepwise logistic regression**