Consensus Class Reference

Generates consensus sequences. More...

#include <Consensus.hpp>

Public Member Functions
	Consensus ()
	Constructor. More...
virtual	~Consensus ()
	Destructor. More...
void	setMissing (char)
	Sets the character interpreted as missing (default: ?)
void	setDisagreement (char)
	Sets the character used to point to disagreements (default: Z)
bool	check_sequences (Align &align)
	Checks all the characters.
Align	consensus (Align &align, char separator='_', bool rigorous=true)
	Reduces the sequence alignment by making consensus sequences. More...
const std::vector< std::string > &	firstSequenceNames ()
	First name of consensed pairs.
const std::vector< std::string > &	secondSequenceNames ()
	Second names of consensed pairs.
const std::vector< std::string > &	roots ()
	Root names of consensed pairs.
const std::vector< int > &	consistentPositions ()
	Number of consistent positions for all consensed pairs.
const std::vector< int > &	complementaryPositions ()
	Number of complementary positions for all consensed pairs.
const std::vector< int > &	uninformativePositions ()
	Number of uninformative positions for all consensed pairs.
const std::vector< int > &	ambiguousPositions ()
	Number of ambiguous positions for all consensed pairs.
const std::vector< int > &	atLeastPartiallyResolvedAmbiguities ()
	Number of at least partially resolved ambiguities for all consensed pairs.
const std::vector< std::vector< int > > &	inconsistentPositions ()
	Vector of inconsistent positions ofr all consensed pairs.

Detailed Description

Generates consensus sequences.

A consensus is generated when two sequences have the same name, ignoring everything after the first separator character (by default, "_"). Hence, the names "foo", "foo_goo" and "foo_third" will be treated as identical and the root will be "foo". The root will be used to name the resulting sequence. Note that the class works only for DNA sequences.

Symbol convention:

• A: adenosine
• C: cytosine
• G: guanine
• T: thymine
• M: A or C
• R: A or G
• W: A or T (weak)
• S: C or G (strong)
• Y: C or T
• K: G or T
• B: C or G or T(not A)
• D: A or G or T (not C)
• H: A or C or T (not G)
• V: A or C or G (not T)
• N: A or C or G or T
• ?: nonsequenced position

Other symbols will be treated as ? (lowercase are supported).

Rigorous (alias liberal or strong) mode:

• If two characters are the same, it is retained whatever it is (A + A = A)

Otherwise:

• If one is the missing character (?) the other is retained whatever it is (A + ? = A).
  • If characters are consistent, that is one contains more information, that one is retained (A + M = A).
  • If characters are not consistent, the closest generic symbol is retained (A + C = M).

Note that the feedback of inconsistent characters in the outcome is not garanteed. In fact, (A + A + G) will result in R (as expected) but (A + G + A) will result in A, masking the problem. However, the position will indeed be counted as inconsistent.

Not rigorous (conservative/weak) mode:

• If two characters are the same, it is retained whatever it is (A + A = A).
• Otherwise:
  • If one is ? the other is retained whatever it is (A + ? = A).
  • Otherwise an inconsistent character (by default, Z) is retained (A + C = Z).

Iterative process of consensus:

• Each sequence is taken in turn.
• Each pair involving the focus sequence is processed and a consensus is generated.
• When all pair have been processsed, the consensus already generated are themselves iteratively processed until only one remains.
• Note that at each time the last two are taken first.

A transparent interface gives access to the data for all steps of the consensus process, as vectors that covers all pairs (including intermediate steps of the iterative procedure described above) as well as singleton sequences. For the latter, the second name is not filled and all counts are set to 0. Note also that the name of such singleton sequence is shortened to the separator as well.

Constructor & Destructor Documentation

Consensus

(

)

Constructor.

virtual ~Consensus ( )

inlinevirtual

Destructor.

Member Function Documentation

Align consensus	(	Align &	align,
		char	separator = '_',
		bool	rigorous = true
	)

Reduces the sequence alignment by making consensus sequences.

Parameters

align	the original alignment.
separator	the character used to separated the root name of sequences to the variable part, as in (for the default value: "sequence_read1".
rigorous	consensus mode.

Returns

An Align instance with duplicated sequences consensed.

---

The documentation for this class was generated from the following files:

• /home/stephane/Documents/devel/egglib2/egglib-cpp/Consensus.hpp
• /home/stephane/Documents/devel/egglib2/egglib-cpp/Consensus.cpp