============================================================================
SOM for RH Panel construction

A new dog/hamster radiation hybrid panel with a higher resolution limit  
has been constructed by fusing dog primary fibroblasts irradiated at
9000 rads with thymidine kinase-deficient hamster cells HTK3 :

I ) Cells:  Dog primary fibroblasts were derived from two male mongrel  
dogs.  Hamster HTK3- cells (thymidine kinase defficient) were derived  
from A2H strain after 6BrdU selection.

II ) Irradiation/Fusion:  A) 20X106 Dog fibroblasts were irradiated  
with 9000 rad.  B) Chemically induced fusion was done using   
PEG2000+10% DMSO with 20X106 HTK3-cells (Dog/Hamster cell ratio = 1.0)

III ) Results: Five identical fusions were done which enabled us to  
produce 400 clones. Average fusion efficiency was 4 per 106 (4 clones  
for every 106 dog fibroblasts).  The 400 clones have been tested for  
their retention frequency using a set of 96 defined BAC markers.  88  
clones were selected and expanded, resulting in a panel with an average  
retention frequency of 29% and a resolution of 300 kb.



- Cell lines retention 
(percentage are derived from 10,500 RH genotyping data).
Cell Line n.1 = 23.0071 %
Cell Line n.2 = 16.3397 %
Cell Line n.3 = 18.6298 %
Cell Line n.4 = 38.3902 %
Cell Line n.5 = 45.3957 %
Cell Line n.6 = 19.1613 %
Cell Line n.7 = 23.8477 %
Cell Line n.8 = 20.0019 %
Cell Line n.9 = 22.0118 %
Cell Line n.10 = 25.2005 %
Cell Line n.11 = 28.4858 %
Cell Line n.12 = 32.0224 %
Cell Line n.13 = 35.7909 %
Cell Line n.14 = 17.7215 %
Cell Line n.15 = 24.5531 %
Cell Line n.16 = 18.4559 %
Cell Line n.17 = 40.0425 %
Cell Line n.18 = 30.7952 %
Cell Line n.19 = 37.8104 %
Cell Line n.20 = 16.7552 %
Cell Line n.21 = 28.4279 %
Cell Line n.22 = 23.5578 %
Cell Line n.23 = 18.1177 %
Cell Line n.24 = 27.8771 %
Cell Line n.25 = 18.0501 %
Cell Line n.26 = 26.0798 %
Cell Line n.27 = 49.3091 %
Cell Line n.28 = 26.7949 %
Cell Line n.29 = 24.2246 %
Cell Line n.30 = 25.4517 %
Cell Line n.31 = 35.4334 %
Cell Line n.32 = 23.2969 %
Cell Line n.33 = 51.0387 %
Cell Line n.34 = 28.5148 %
Cell Line n.35 = 25.674 %
Cell Line n.36 = 37.5882 %
Cell Line n.37 = 43.5791 %
Cell Line n.38 = 47.8404 %
Cell Line n.39 = 41.8688 %
Cell Line n.40 = 27.0268 %
Cell Line n.41 = 15.3445 %
Cell Line n.42 = 25.7996 %
Cell Line n.43 = 37.965 %
Cell Line n.44 = 23.5965 %
Cell Line n.45 = 23.7318 %
Cell Line n.46 = 43.2022 %
Cell Line n.47 = 23.21 %
Cell Line n.48 = 16.533 %
Cell Line n.49 = 22.2534 %
Cell Line n.50 = 48.2655 %
Cell Line n.51 = 18.5718 %
Cell Line n.52 = 10.5904 %
Cell Line n.53 = 33.5395 %
Cell Line n.54 = 38.1486 %
Cell Line n.55 = 22.5722 %
Cell Line n.56 = 38.738 %
Cell Line n.57 = 21.1711 %
Cell Line n.58 = 30.1092 %
Cell Line n.59 = 15.0449 %
Cell Line n.60 = 21.9152 %
Cell Line n.61 = 27.1041 %
Cell Line n.62 = 22.0118 %
Cell Line n.63 = 20.6687 %
Cell Line n.64 = 34.0613 %
Cell Line n.65 = 19.1323 %
Cell Line n.66 = 23.3936 %
Cell Line n.67 = 21.7992 %
Cell Line n.68 = 20.8909 %
Cell Line n.69 = 30.5827 %
Cell Line n.70 = 17.7215 %
Cell Line n.71 = 29.3458 %
Cell Line n.72 = 20.3015 %
Cell Line n.73 = 44.4487 %
Cell Line n.74 = 16.2335 %
Cell Line n.75 = 35.2401 %
Cell Line n.76 = 30.3701 %
Cell Line n.77 = 30.1962 %
Cell Line n.78 = 32.0128 %
Cell Line n.79 = 13.9434 %
Cell Line n.80 = 38.2452 %
Cell Line n.81 = 21.9152 %
Cell Line n.82 = 41.2407 %
Cell Line n.83 = 22.1084 %
Cell Line n.84 = 27.7708 %
Cell Line n.85 = 23.9443 %
Cell Line n.86 = 32.3799 %
Cell Line n.87 = 20.6977 %
Cell Line n.88 = 37.0374 %

RH panel average retention : 29 % 

Histogram of cell line retention [-full size-] 


====================================================================================
SOM for Markers identification

All canine sequence were identified through BLASTn best mutual  
alignement with the human genome.Human genes that had a representative  
dog sequence were allotted to 11818 bins . The human genome was split  
into 75 kb segments and the sequence providing the highest blastn score  
in each of the 11818 binsAlignment was retained whenthe dog/human  
alignment overlaps at least one human exon defined by the 24,567 genes  
of Ensembl version 13.31.1

All selected dog sequences are contigs or singletons, derived from the  
assembly of 6.22 million reads. All dog sequence were searched against  
the human genome (NCBI 31) by blastn.

Selection criteria :

1. Blastn alignment with human genome yields mutually-best HSPs.

   HSPs are defined as mutually-best when the segment of  
aligned dog sequence has no higher-scoring hit elsewhere on the human  
genome AND the segment of aligned human sequence has no higher-scoring  
alignment with another dog sequence.

2. The dog/human alignment overlaps at least one human exon

           Defined by the 24,567 genes of Ensembl version 13.31.1

3. The dog sequence is not a processed pseudogene

           Where multiple HSPs were separated by <25 bases on a dog  
sequence, but >300 bases on the human sequence, the dog sequence was  
considered as a potential pseudogene, and was eliminated.

4. The dog/human alignment includes only part of the dog sequence

           A dog sequence for which the alignment extended over the  
complete length of the dog sequence may not yield dog-specific primers,  
and was eliminated.

5. After applying criteria 1-4, selection of a single dog sequence for  
each human gene    The sequence providing the highest blastn score.

6. Human genes that have a representative dog sequence were allotted to  
11818 bins

           i.e. The human genome was split into 75 kb segments. If  
multiple human genes fell within a single segment, these were binned  
together.

7. Selection of a single dog sequence for each 75 kb bin

           The sequence providing the highest blastn score in each of  
the 11818 bins.

8. Selection of a dog sequence from alternate bins

           This gave 5909 sequences for the first round of mapping.  
There remains another 5909 dog sequences from the remaining 75 kb bins  
that can be used for the second round.

 

================================================================================
SOM for Map construction

Map construction was carried out with the rh_tsp_map2 package (Agarwala  
et al. 2001) that uses the well-known state of art CONCORDE algorithm  
in the TSP problem (Applegate et al.) The RH vectors were analysed by  
pairwise analysis using the pairlods_dists option to constitute linkage  
groups. The Traveling Salesman Problem (TSP) method using the CONCORDE  
chained Lin-Kernighan algorithm of the rh_tsp_map2 package was used to  
order of each linkage group. Parametric and non-parametric analysis is  
systematically provided resulting in 5 independent maps. A consensus  
map could be derived from the 5 independent maps along with a mapping  
support expressed in percentage assigned to each marker.





Legend :


The rh_tsp_map2 program delivers 5 maps. 
Results of the comparison of these five maps 
is enlighted by horizontal bars. 
When the same marker is present in the 
five maps at a given position the 
horizontal bars have a maximum length 
and correspond to mapping positions 
reaching high confidence ( i.e. box 1). 
Map positions occupied by two or more 
markers are characterized by shorter horizontal 
bars (i.e. boxes 2 and 3). 
Although as in box 2 a given marker can 
occupied three different positions 
extending the size of the scrambled 
region, in box 3 only two adjacent 
markers exchange their positions 
limiting the zone of uncertainty


 from Hitte et al. 2003





















































================================================================================
SOM : MAP RESOURCES 

 ALL RH vectors are available : [ -download vectors- ]

 ALL gene sequences are available  : [ -download sequences- ]
 
Unix users :
1: download the file 
2: gunzip ALL-seq.tar.gz # uncompress the file 
3: tar xf ALL-seq.tar    # each chromosome are distributed in individual tar ball. 
4: tar xf CFAxx.tar      # uncompress individual chromosome files

 The High resolutive canine RH panel is available upon request :  [- ask for RH panel -]