The XML output of mga consists of a summary of this particular run and of the computed partial multiple alignment.
<!ELEMENT MgaOutput (Summary, PartialMultipleAlignment)>
The summary itself is subdivided into the parameters mga was called with, the description or names of the sequences as found in their FASTA headers, other information about the sequences and finally some numbers.
<!ELEMENT Summary (CommandLineArguments, SequenceDescription+,
SequenceInformation+, TotalNumber+)>
The command line arguments are some sort of character data and not structured into subunits any further.
<!ELEMENT CommandLineArguments (#PCDATA)>
The sequence names are also unstructured.
<!ELEMENT SequenceDescription (#PCDATA)>
The additional information is comprised of several lengths.
<!ELEMENT SequenceInformation (Length+)>
The total numbers and the lengths are natural numbers.
<!ELEMENT TotalNumber (#PCDATA)>
<!ELEMENT Length (#PCDATA)>
A partial multiple alignment consists of at least one block. An extreme case is a single unaligned block, i.e., all sequences are in fact unaligned.
<!ELEMENT PartialMultipleAlignment (Block+)>
A block is comprised of all sequences which means at least two sequences, or else an alignment would not make sense.
<!ELEMENT Block (OneSequence+)>
A particular sequence (fragment) carries region information describing where it resides on the genome. Furthermore, its sequence data is stored.
<!ELEMENT OneSequence (Region, SequenceData)>
A region simply consists of a start and an end point.
<!ELEMENT Region (Start, End)>
Sequence data can contain any kind of character as mga per se does not restrict the alphabet to consist of A, C, G, T only.
<!ELEMENT SequenceData (#PCDATA)>
Start and end points are natural numbers.
<!ELEMENT Start (#PCDATA)>
<!ELEMENT End (#PCDATA)>