CRI-MAP Tutorial - Data File Formatting

---

CRI-MAP tutorial contents:

Manuals: Web & text versions	Web Manual Table of Contents	Tutorial   Practice   Datasets	Mapping with "build"	Testing & Extending Maps	Bibliography & Other Links

1. The initial data file, chr#.gen

Ex. 1 - create an input file

2. Auxilliary data files, chr#.loc   chr#.ord   chr#.par   chr#.dat  (prepare option)

Ex. 2 - prepare the auxilliary data files

3. Merging data files with common loci or families  (merge option)

Ex. 3.1 - merge two chr#.gen files

Ex. 3.2 - prepare new auxilliary data files

1. The initial data file - chr#.gen       (... & here's what the manual says)  

You must supply the initial data file to CRI-MAP. It is a plain "ascii" text file that may be created and edited in almost any word-processing programme. chr#.gen data files always have the form:

(Section 1, with info for the entire pedigree collection:)

# of families
# of loci
locus_name1  ln2  ln3 etc.    (NB: locus names have a 15 character limit)

(Section 2, with the pedigrees:)
    (Section 2.1-f, with general info for the first family:)

   family 1 id                (any unique name or number designation)
   # of family members

        (Section 2.1-i, with the data for each individual from the first family:)

       individual 1 id#   mother's id#   father's id#   sex id#
       locus_name1allele1#  ln1a2#  ln2a1#  ln2a2# etc.
       individual 2 id#   mother's id#   father's id#   sex id#
       ln1allele1#  ln1a2#  ln2a1#  ln2a2# etc.
       etc. 

   family 2 id
   # of family members

        (Section 2.2-i, with the data for each individual from the second family:)

       individual 1 id#   mother's id#   father's id#   sex id#
       locus_name1allele1#  ln1a2#  ln2a1#  ln2a2# etc.
       individual 2 id#   mother's id#   father's id#   sex id#
       ln1allele1#  ln1a2#  ln2a1#  ln2a2# etc.
       etc.
 
   and so on ...

The manual shows this short generic example.  

Points to note:

• locus names are separated by spaces, so any particular locus name may NOT contain a space character
• locus names and family names may be any combination of letters, punctuation, and/or numbers
• "id#"s are integers - ALL alleles must be coded as integers
• for individuals without parents in the pedigree, "mother's id# & father's id#" are coded as 0
• for "sex id#", female=0 male=1 unknown=3
• unknown "allele#"s are coded as 0
• for X-linked loci, code the non-existent second alleles in males (for the Y-chromosome) with an unused ("dummy") allele number

Exercise CRI-MAP 1: create an input chr#.gen file

Using this key to map allele names to allele numbers, code the following pair of pedigrees into an input file for CRI-MAP. Use a word processor that allows file saving in text format. Name the file chr37.gen.

Chromosome 37 Loci
Locus Name     allele (code #)     allele (code #)     allele (code #)
     A            a      (1)          A      (2)
     B            b      (1)          B      (2)
     C            c      (1)          C      (2)          c*     (3)
     D            d      (1)          D      (2)          d*     (3)
     E            e      (1)          E      (2)

Family LMN.3 Chr. 37 Loci         Family OPQ.7 Chr. 37 Loci

NB: When using your own data, if they are already in the correct format for analysis by the Linkage suite of programmes, you may convert them to CRI-MAP chr#.gen format with the LNKTOCRI utility. 

2. The auxilliary data files - via the prepare option

• chr#.loc
• chr#.ord      
• chr#.par      
• chr#.dat      

With the chr#.gen file written, you submit it to the prepare option of CRI-MAP to create standard versions of the auxilliary data files. ALL CRI-MAP options are applied to datafiles with commands of this form:

prompt> crimap # option

For example, to perform "twopoint" linkage analysis on data for chromosome 12 (you will already have datafiles named chr12.gen, chr12.dat, chr12.loc, chr12.ord, & chr12.par), the command is:

prompt> crimap 12 twopoint

The programme then responds by displaying the results of the analysis on the screen. Unlike most other CRI-MAP options, however, prepare is an interactive one; before it completes the preparation of the auxilliary datafiles, you must reply to a series of questions, mostly having to do with accepting default values for various parameters and filenames used in the analytical options.

Exercise CRI-MAP 2: prepare the auxiliary data files

If you haven't already done so, copy your plain text chr37.gen file into a new directory in your filespace at your EMBnet Node host computer. Name the directory Crimaptutl, and switch to it when ready to proceed. (Unfamiliar with UNIX? Check the UNIX Commands Short List page.)

Enter the command crimap 37 prepare at the prompt, and accept the default values or enter the answers shown in the series of questions that follows.

• When asked " Do you wish to enter any new haplotyped systems?", enter n.
• When asked " Do you wish to hold any additional recombination fractions fixed?", enter n.
• Choose build as the next option to run, by entering "1".
• When asked " Do you wish to build the map incorporating ALL loci in decreasing order of their informativeness (the usual procedure)?, enter y.
• And finally, when asked " OK to set up new parameter file?, enter y.

NB: If you see the warning:
NONINHERITANCE: family "fam_name", individ id#, locus #
your chr37.gen file has an error!

Although we have chosen build as the next option to run, first examine the structure and the content of the datafiles just created from the chromosome 37 pedigrees. These datafiles summarise different aspects of the data and details of the subsequent processing by CRIMAP.

chr37.loc is a file holding only information on the loci.

Quoted below is the chr37.loc file derived from only Family LMN.3; it shows the number of loci, the index numbers and names of each, plus two columns for "informative meioses".

Genotype file chr37lmn.gen

Number of loci:  5

                       #inf. mei.    #inf. mei.(phase known)
  0  A                      3              3
  1  B                      3              3
  2  C                      7              3
  3  D                      11             0
  4  E                      4              0

Traditionally, an informative meiosis is one that, firstly, leads to offspring, and secondly, occurs in a parent who is (usually) at least a double heterozygote. In Family LMN.3, the grandfather is a triple heterozygote who produces four children. Thus, he yields four informative meioses (one per child) for each of the three loci: C, D, & E. His son who is a triple heterozygote (loci A, B & C) produces three children, thus yielding three informative meioses for each of these three loci. Summing up, the totals are: A-3, B-3, C-7, D-4, & E-4. Why then has locus D eleven informative meioses, according to CRI-MAP? The prepare option has determined that both the grandmother and the son's partner - even though they are only single heterozygotes - also provide informative meioses for the D locus. There are four meioses in the grandmother (her four children) plus three more from her son's partner. This adds the extra seven informative meioses for locus D.
Note that the informative meioses from the son are also phase known. This means that we know which set of five alleles for these five loci came from each parent. We know the original allele complements (phase) of his two chromosomes. He received a chromosome with alleles A b c* d E from the grandmother, and another with alleles a B C d E from the grandfather. Knowing the phase of the chromosomes reduces the number of possible original-states the programme must explore in its analysis, thus speeding result and increasing its accuracy.

chr37.ord is the file holding the initial information on locus order.

Quoted below is the initial chr37.ord file produced by prepare.

1


1  2
  2   0  

For this simple chromosome 37 example, the prepare option recognises only one set of possible orders, and this set has only one member, the trivial order consisting of the two most informative loci. The order itself (trivial because it has only two loci) is shown via the index numbers of the loci: "2   0".
As the chromosome map is estimated by CRI-MAPs build option, the chr37.ord file will be modified and updated, finally reporting ALL likely locus orders. A "likely" locus order is one within a given range of the "most likely" locus order(s), the one(s) with the lowest LOD score. We'll examine this file again after using the build option.

chr37.par contains the values for all the parametres used by the various CRI-MAP options.

The values for the many parametres are the default ones accepted during the prepare operation. Detailed descriptions of the parametres, their default values, and how they are used by the CRI-MAP options are in the manual.
For now, notice that the chr37.par file holds the names of the other auxilliary data files,
    dat_file chr37.dat *
    gen_file chr37.gen *
    ord_file chr37.ord *
tells not to use the chr37.ord file in the subsequent build option
    use_ord_file 0 *
though it may be altered
    write_ord_file 1 *
and shows the index numbers of the loci having an established order
    ordered_loci 2 0 *
and of those waiting to be established.
    inserted_loci 1 3 4 *

chr37.dat shows the phase known data available to the build option.

The phase known information is organised by chromosome and by family in this file, as opposed to by locus in chr37.loc.

3. Merging data files with the merge option

As you acquire new data, you may add it to existing data files automatically. merge allows the addition of data when either loci or families are shared between separate chr#.gen files.

For example, we could increase the amount of phase known information for the chromosome 37 dataset if we knew the genotypes of the parents of the two males who "bred into" family OPQ.7 .

Exercise CRI-MAP 3.1: merge two chr#.gen files

Copy the plain text chr37OPQ.gen file into your Crimaptutl, directory.

Enter the command crimap 37 merge at the prompt, and enter the answers shown in the series of questions that follows.

• When prompted for " first input file =", enter chr37.gen.
• When asked " second input file =", enter chr37OPQ.gen.
• When asked " output file =, enter chr37a.gen.

NB:You will see warnings similar to:

 MISMATCH IN PEDIGREE 537728400, individual 7 

 MISMATCH IN PEDIGREE 537728400, individual 8 

The chr37OPQ.gen file has no error - these two individuals are the two males who have just had their parents added to the pedigree. As such, their "mother's id#" & "father's id#" values change from "0" (unknown) to one of the four new pedigree members.

What additional information is now available to CRI-MAP? A quick way to check is to compare the chr37.loc file with the new chr37a.loc file you can create via the crimap 37a prepare command.

Exercise CRI-MAP 3.2: prepare new auxilliary data files using the added data

Enter the command crimap 37a prepare at the prompt, and accept the default values or enter the answers as before.

Note that locus E is now reported as being more informative than locus D by the prepare option output. Comparison of chr37.loc with chr37a.loc shows four more informative meioses overall, 10 more being phase known, and three of these new phase known meioses occurring at locus E.

Please continue with the build, twopoint, flipsn, instant, & fixed options, in
Part 2 - Mapping & LOD scores