Wheat Pedigree

From ICISWiki

Jump to: navigation, search

Annex 11. Definition of a Pedigree

Wheat Pedigree Management

Many of the questions asked by wheat breeders fall intro two categories: Which lines are related to X? What information do we have about line X? Both of these questions imply that X can be clearly and unambiguously identified. Given the way in which wheat pedigrees are identified this is not always easily determined.

The pedigree notation used by CIMMYT and many other breeding institutions is based on what is commonly known as the Purdy method (Crop Science, Vol. 8, July-August 1968). This notation gives an explicit description of the crossing and selection history of a line. Such descriptions tend to become rather long after a few generations of crossing and are therefore given an abbreviated name. The problem then becomes one of recognizing and remembering what the abbreviations stand for. The problem is further compounded when several sister lines are given different abbreviations. Another source of confusion results from reassignment of names by commercial seed companies and national breeding programs.

The limitations described above should not detract from the usefulness of the method for it has withstood the test of time and many organizations still use it. What is needed is a structure for managing the complexity of pedigrees and their relationships. This we define as Pedigree Management. The following is a functional description of the Pedigree Management of IWIS.

Definition of a Pedigree

For the purpose of Pedigree Management, a pedigree is defined as: a cross name plus a selection history. These terms are more fully defined in later. However, be aware that this definition is somewhat different from what is normally meant by the word "pedigree" at CIMMYT. The word "pedigree" as it is often used at CIMMYT is referred to as a selection history in this document.

IWIS makes certain assumptions concerning the syntactical analysis and semantic interpretation of pedigrees. It is important to understand that any pedigrees that do not fully comply with the rules described in the following sections cannot be included in any of the databases used by IWIS (including the field book system).

The following defines the terms relating to Cross Names as they are used throughout this document.

CROSS NAME. A cross name is the name given to a cross. The cross name may be a single name (simple cross name) or it may be a composite name containing previously defined simple cross names, each of which is separated from other simple cross names by a cross symbol.

SIMPLE CROSS NAME. A cross name that does not contain cross symbols. Simple cross names are names that are given to land races, breeding lines and cultivars. In general, a simple cross name is a synonym for a composite cross name - selection history combination. It may also be a synonym for a breeding line, that is, a cross and all of its progeny (siblings).

COMPOSITE CROSS NAME. A cross name containing two or more simple cross names each separated by a cross symbol. Within a composite cross name the female parent is always displayed to the left of the cross symbol and the male parent to the right (each parent may be identified by a simple or composite cross name).

CROSS SYMBOL. Cross symbols are used to separate simple cross names within a composite cross name. The cross symbol indicates either a simple cross or a back cross. Taken as a group the cross symbols in a composite cross name explicitly indicate the order in which crosses were made.

SIMPLE CROSS SYMBOL. Simple cross symbols are the slash, ’/’, indicating a first order cross; double slash marks; ’//’, indicating a second order cross; or a single slash mark followed by a natural number ’n’ and a second slash mark. The number ’n’ indicates a cross of order ’n’ and is used for third order crosses and higher.

BACK CROSS SYMBOL. Back cross symbols are composed of an asterisk, ’*’, a natural number and a simple cross symbol. The asterisk separates the name of the back cross parent from the natural number. The natural number identifies the number of times the back cross parent was used. If the asterisk is on the left side of the simple cross symbol, then the back cross parent is the female. If the asterisk is on the right side of the simple cross symbol, then the back cross parent is male. A back cross symbol contains exactly one asterisk.

The following defines the terms relating to Selection Histories as they are used throughout this document.

STANDARD SELECTION HISTORY. A standard selection history conforms to the rules described under the sections entitled "Selection History Syntax" (page 6) and "Selection History Semantics" (page 7). The term "selection history" implies a standard selection history unless stated otherwise. All selection histories generated by IWIS will conform to the standard selection history syntax and semantics. A standard selection history consists of the following items: A crop identifier; a cross number; the year of crossing; an optional cross type identifier and plant selection identifiers. Each item is separated from the others by a hyphen.

NON-STANDARD SELECTION HISTORY. A non-standard selection history is any conglomeration of characters that uniquely identify a particular cross and the selections made from the progeny of the cross. IWIS offers limited support for non-standard selection histories. However, non-standard selection histories convey no syntactic or semantic information other than to identify germplasm as being unique.

CROP IDENTIFIER. The crop identifier is a one-to-three letter code that identifies the type of crop and/or location where the cross was made (e.g. CM, CWH, etc.).

CROSS NUMBER. A five digit natural number that individually or taken together with the year of crossing uniquely identifies the cross name to which the plant selection history applies.

CROSS YEAR. The year of crossing is a two digit number that identifies the year the cross was made (e.g. 85 for 1985).

CROSS TYPE. The cross type is an optional term in a standard selection history that identifies if the cross under consideration is a simple (S), back (B), top (T), double (D) or wide cross (W). A second letter or digit may be used to indicate uniqueness of the cross (i.e. if the same top or double cross was performed multiple times).

PLANT SELECTION IDENTIFIERS. The plant selection identifiers consist of a number followed by one-to-three letters and an optional field delimited by parentheses. The optional field is used to indicate from what nursery the selection was made (e.g. drought screening). The number indicates how many plants were selected and/or the selection method. The letters indicate where the selection was made. Three selection types are recognized.

  • Single plant selection. The selection number is greater than 0 and less than 500 and does not contain leading zeros. The cardinality of the number indicates the relative selection made from the row (e.g. 23 indicates the 23rd plant to be selected from the row). Each selection is then grown as a single row in the next generation.
  • Modified bulk. The selection number has a leading zero. The cardinality of the selection number indicates how many plants were bulked from a single row. The bulked plants are then grown as a single row in the next generation.
  • Bulk. The selection number is zero indicating a bulk of the entire row.

Cross Name Syntax

The following is a formal syntactical description of cross names. The definition is given in Backus-Naur Form and constitutes a context-free grammar.

CROSS_NAME ::= <name>
    <name><symbol><cross_name>
SYMBOL ::= <simple_cross>
    <back_cross>
SIMPLE_CROSS ::= /
    //
    /<integer>
BACK_CROSS ::= *<integer><simple_cross>
    <simple_cross><integer>*
NAME ::= <character>
    <character><name>
INTEGER ::= <digit>
    <digit><integer>
CHARACTER ::= <digit>
    <letter>

Examples of syntactically valid cross names:

SON 64/KLRE
SON 64//6*SEE/3*ANE
KL.H686.F2600//SR/TPB/CNO"S"
CTFN/8156(R)/3/CHR//TT/SON
64/4/FKN

Examples of syntactically invalid cross names:

TOB"S"*68/CIAN0 Back cross level greater than 60
CNO/BOW"S"/ Cannot begin or end cross names with a cross symbol
A*B/SON 64 Name cannot contain ’*’ or ’/’
VEE"S"*2/2*SON 64 Cross symbol contains more than 1 asterisk

Cross Name Semantics

The semantic interpretation of a cross name involves determining the parental role (either male or female) and the order in which crosses were made.

The semantics of a cross name can easily be described in the form of a binary tree where the root of the tree represents the latest cross. The left sub-tree represents the female parent and the right sub-tree represents the male parent. The tree can be constructed using the following rules:

  • Find the highest order cross symbol for the portion of the cross name under consideration.
  • Construct a node where the left branch points to the female parent and the right branch points to the male parent. The parents may themselves be composite crosses.
  • If the current cross is a back cross, repeat the following step ’n-1’ times where ’n’ is the back cross level:
  • For recurrent male parent: Replace the female parent with a new node having branches pointing to the original female parent and the male parent.
  • For recurrent female parent: Replace the male parent with a new node having branches pointing to the male parent and original female parent.
  • Consider the new node for the next (if any) replacement operation.
  • If the branch pointing to the female parent has not been decomposed into its composite crosses then decompose it using the rules described above. Otherwise, go back to the previous node and consider the male branch from that node. If the male branch has not been decomposed, then decompose it using the rules described above.

The process is complete when all nodes of the tree have been processed, that is, when there is no previous node to go back up to.

The following family trees were constructed by applying the above set of rules.

A/B/3/C/D//E

┌──────────┐
│ │
┌──┴──┐ ┌─┴──┐
│ │ │ │
A B ┌─┴─┐ E
│ │
C D
A*2/B//C

┌──────┐
│ │
┌──┴──┐ C
│ │
A ┌─┴─┐
│ │
A B

Examples of syntactically valid but semantically invalid cross names

A/BIG Order of crossing not determinable
A/4/BIG/ID Third order cross level missing

Selection History Syntax

The following is a formal syntactical description of a selection history. The definition is given in Backus-Naur Form and constitutes a context-free grammar.

SELECTION HISTORY ::= <cross_id><cross_type><selections>
CROSS_ID ::= <program>-<cross_number>-<cross_year>
PROGRAM ::= <letters>
CROSS_TYPE ::= -T<letters>
    -D<letters>
    -S
    -W
    -B
SELECTIONS ::= Null
    -<digits><letters><nursery><selections>
NURSERY ::= Null
    <letters>
CROSS NUMBER ::= <digit><digit><digit><digit><digit>
CROSS YEAR ::= <digit><digit>
LETTERS ::= <letter>
<letter><letters>
DIGITS ::= <digit>
    <digit><digits>

Examples of some syntactically valid selection histories:

CM-00783-85-TA

CMW-98329-84-1Y-5M(DR)-0100Y-20M-OY

CD-52761-85-DAA-2B-1Y


Examples of some syntactically invalid selection histories:

CMW1-98732-84 Crop type is too long and contains digits
CM-1234-85-10Y Cross number must be five digits
CM-12345-10Y Cross year is missing
CD-00102-84-4B-90PTZA Too many letters in selection

Selection History Semantics

The semantic interpretation of selection histories is relatively straight forward. The cross identifier should be unique for any given cross name (an exception is introduced in the section "Cases of Non-Unique Selection Histories" – page 7). The crop type determines the type of cereal that was involved in the cross, e.g. bread wheat. The cross number is a number identifying an individual cross for the specified crop and year. The year is the year the cross was made. The cross type, if present, identifies the cross as being a simple, wide, top or double cross. Any letter(s) that follow the cross type indicate a specific top or double cross made using the same parents. The absence of the cross type implies that a simple cross was made. The selections identify the nursery, location and type of selection made from a segregating population.

The order and number of the selection terms is important. The first term identifies the selection made from the Fl generation. Later terms identify later filial generations and the selections made on them. In general, the selections are made in successive generations. However, there are exceptions, particularly in the case of re-selections.

Relationship Between Cross Name and Selection history

Normally a selection history uniquely identifies a cross and a particular set of selections made on the progeny of the cross. The selection history does not explicitly indicate the parents involved in making the cross.

In general, the cross name that corresponds to the cross identifier part of a selection history identifies parents by name and order of crossing to some arbitrary level. However, none of the selection history is conveyed by the cross name itself.

There is a one-to-many relationship between cross names and selection histories. In other words, each cross name may have associated with it several selection histories. A selection history will have exactly one cross name associated with it, (except in certain special cases, see below).

Cases of Non-Unique Selection Histories

Some exceptions to the above are recognized. A plant may be found in the field. Nothing is known about the plant’s pedigree but it is introduced into the breeding program because of one or more exceptional qualities. These plants are assigned an arbitrary but unique cross name. A selection history of the following form is also assigned:

<crop_ id>-00000-<cross_ year>

In this case the <cross_ year> is the year in which the plant was found. Further selections may be made in subsequent cycles.

There are also cases where a named cultivar or breeding line is used in a breeding program but its selection history is not fully known. Such material is assigned a non-standard selection history where the identifying string contains as much information as is known. This may vary from nothing to nearly complete information. Neither the selection history nor the cross name need be unique in themselves. However, the combination of cross name and selection history must be unique.

The above case includes non-CIMMYT material where the pedigree does not conform, and cannot be made to conform, to CIMMYT standards.

Examples of Cross Name and Selection History Usage

Consider the following cross name: CN067/BB"S". The cross name indicates that a cross was made with CNO67 as the female parent and BB"S" as the pollen source. There will be several selection histories corresponding to this cross name. Each of the associated selection histories will have the same cross identifier (see CROSS_ ID in the BACKUS-NAUR FORM description of Selection History Syntax – page 6).

Reference must be made to the selection history associated with the cross in order to uniquely identify any of the selections made from the progeny of this cross. In addition, it may be necessary to make reference to the selection history of the parent, BB"S", in order to properly identify exactly which sibling from the breeding line called BB was used in making the original cross. This becomes particularly important to breeders if they wish to make a backcross with BB"S" as the recurrent parent.

Consider three siblings of BB"S":

II23584-14Y-1M-9Y-2M sibling 1
II23584-12Y-1M-9Y-3M-OY sibling 2
II23584-12Y-1M-9Y-2M sibling 3

Now suppose that the original cross was made with sibling l. If we crossed sibling 1, as the pollen source, with CNO67/BB"S", the resulting cross name would be:

CNO67/2*BB"S"

However, if the cross was made with sibling 2 as the pollen source the resulting cross name would not indicate a backcross, rather, it should be:

CNO67/BB"S"//BB"S"

This cross name explicitly excludes the possibility that the same BB siblings were used. At the same time, it does not give any indication which siblings were actually used.

The only way that the proper cross name can be constructed is through reference to the selection histories of the parents involved in making the cross.

Relationships

Cross names and selection histories contain considerable information with respect to determining how closely ’related’ various lines and cultivars are.

Intra-Line Relationships

Selection histories may be used to determine intra-line relationships. Consider the selection histories for the three siblings of BB illustrated above. It can be seen that sibling 1 has a very different selection history from siblings 2 and 3 (they are different selections from Fl onward). This would indicate that sibling 1 has fewer genes in common with sibling 2 than sibling 2 does with sibling 3.

Inter-Line Relationships

Cross names may be used to determine more general relationships between breeding lines and cultivars. Considering the cross between CNO67 and BB"S". Again we may assume that the genes held by such a cross would he contributed on a 50:50 basis by the parents. By similar reasoning one would assume that by making the back cross with BB"S" the resulting offspring would potentially contain 25 percent of the genes from CNO67 and 75 percent of the genes from BR"S". These ratios assume Mendelian genetics.

Now consider the expanded cross names of each parent:

BB"S" = CNO67"S"//SON64/KLRE/3/8156
CNO67 = PI/CHR"S"//SON64

We can see that BB"S" itself is a cross involving a sibling of CNO67. Thus it should become clear that the genetic contribution of CNO67 in any cross with BB"S" is relatively high. Another complicating factor in this cross is that BB"S" and CNO67 both contain a common parent, SON64.

Cross names may be further expanded in order to determine what additional relationships may exist.

Given proper tools to examine both inter- and intra-line relationships, breeders may be able to determine where desirable genes are ’coming from’ or where to begin searching for them.

The number of selections made on the progeny of crosses may be used to obtain a rough idea of the relative combining ability of certain commonly crossed lines and cultivars. The principle being that if may selections are made from a specific cross one or both of the parents are good combiners. On the other hand if few or no selections are made one or both of the parents are poor combiners.

Naming of Crosses

Cross names can become very long and cumbersome to use, especially when the cross is frequently used to make more crosses. The length of the resulting cross name then becomes the sum of the lengths of the two parents.

For example consider the cross:

BY/MAYA"S"/4/BB//HD832.5.5/ON/3/CNO/PJ

If we were to cross a sibling from this line with another having a name of equal length the resulting name would be unmanageable.

In order to make life a little easier, breeding lines are assigned names or pseudonyms. The above cross name was assigned the name BUCKBUCK. Even this name is a bit long and therefore it may be further abbreviated to simply BUC. Since the name BUC refers to the cross and not a specific selection from that cross (i.e. BUC does not imply a specific sibling or cultivar) it has been customary to append an "S" to the name.

Names of lines vs. Names of Cultivars

It is not uncommon for a breeding line or cultivar to be known by several different cross names. This is particularly common among very successful crosses. An example is the breeding line called VEERY. Any of the siblings of the original VEERY cross may known as VEERY"S" or by the short form name VEE"S".

Of the VEERY siblings, some have been given variety names. For example, the sibling identified by the selection history CM-03327-81-TF-15M-500Y-OM is known as VEERY_ 5. It is also known in various parts of the world as SERI 82, PAKISTAN 81, SERIC, plus a few more names. All of these names refer to the same pedigree, that is VEERY CM-03327-8l-TF-15M-500Y-0M.

A second, but different, selection of VEERY is known as VEERY_ 3, LIMA 1 or GAMTOOS. GAMTOOS and SERIC are both VEERY siblings but different selections, and therefore, different pedigrees.

These relationships are not immediately obvious given only cross names. Furthermore, it is not possible to determine whether a name is a reference to a breeding line (many different selections) or a cultivar (a single selection).

Personal tools