Previous  Up  Next

Module Biocaml_fastq (.ml)

FASTQ files. The FASTQ file format is repeated sequence of 4 lines: 

    @name
    sequence
    +comment
    qualities
    ...

The name line begins with an @ character, which is omitted in the parsed Biocaml_fastq.item type provided by this module. Any spaces after the @ are retained, but the specification implies that there shouldn't be any such spaces. Trailing whitespace is also retained since you should not normally have such files.

The comment line, which begins with a +, is handled similarly. The purpose of the comment line is unclear and it is rarely used. Also, "comment" may not be the correct term for this line.

The name line may be structured into two parts: a sequence identifier and an optional description. We provide a function Biocaml_fastq.split_name to parse such a value. However, an item's name field contains the unparsed string because it is unclear whether fastq files really follow this. Also the format of the description is unspecified. When it is provided, usually it has some additional structure, so the minimal amount of parsing done by Biocaml_fastq.split_name isn't too useful anyway.

Illumina uses a systematic format for the name line that serves as a unique sequence identifier. Use Biocaml_fastq.Illumina.sequence_id_of_string to parse an item's name field when you have fastq files produced by Casava version >= 1.8. Earlier versions of Casava returned a different format, which is not currently supported in this module (it could be easily added).

The qualities line is returned as a plain string, but it is required to be decodable as either Phred or Solexa scores. Modules Phred_score and Solexa_score can be used to parse as needed.

Older FASTQ files allowed the sequence and qualities strings to span multiple lines. This is discouraged and is not supported by this module.

module Biocaml_fastq:
sig
type item = {
   name : string;
   sequence : string;
   comment : string;
   qualities : string;
} 
val split_name : string -> string * string option
Split a name string into a sequence identifier and an optional description. It is assumed that the given string is from an item's name field, i.e. that it doesn't contain a leading @ char.

Input/Output 


module MakeIO: 
functor (Future : Future.S) ->
sig
val read : Future.Reader.t ->
       Biocaml_fastq.item Core.Std.Or_error.t Future.Pipe.Reader.t
val write : Future.Writer.t ->
       Biocaml_fastq.item Future.Pipe.Reader.t -> unit Future.Deferred.t
val write_file : ?perm:int ->
       ?append:bool ->
       string -> Biocaml_fastq.item Future.Pipe.Reader.t -> unit Future.Deferred.t
end
include ??

Illumina-specific operations


module Illumina:
sig
type surface = [ `Bottom | `Top ] 
type tile = private {
   surface : surface;
   swath : int; (*1, 2, or 3*)
   number : int; (*1 - 99, but usually 1 - 8*)
} 
val tile_of_string : string -> tile Core.Std.Or_error.t
E.g. tile_of_string "2304" parses to
  • surface = Bottom
  • swath = 3
  • tile_num = 4

val tile_to_string : tile -> string
Inverse of tile_of_string.
type sequence_id = private {
   instrument : string;
   run_number : int;
   flowcell_id : string;
   lane : int;
   tile : tile;
   x_pos : int;
   y_pos : int;
   read : int;
   is_filtered : bool;
   control_number : int;
   index : string;
}
Content of name lines as generated by Casava versions >= 1.8.
val sequence_id_of_string : string -> sequence_id Core.Std.Or_error.t
Parse a name string to a structured Illumina sequence_id. It is assumed that the given string is from an item's name field, i.e. that it doesn't contain a leading @ char.
end

Low-level Printing


val item_to_string : item -> string
This function converts item values to strings that can be dumped to a file, i.e. they contain full-lines, including all end-of-line characters.
val name_of_line : ?pos:Biocaml_internal_utils.Pos.t ->
       Biocaml_internal_utils.Line.t -> string Core.Std.Or_error.t

Low-level Parsing


val sequence_of_line : ?pos:Biocaml_internal_utils.Pos.t -> Biocaml_internal_utils.Line.t -> string
val comment_of_line : ?pos:Biocaml_internal_utils.Pos.t ->
       Biocaml_internal_utils.Line.t -> string Core.Std.Or_error.t
val qualities_of_line : ?pos:Biocaml_internal_utils.Pos.t ->
       ?sequence:string ->
       Biocaml_internal_utils.Line.t -> string Core.Std.Or_error.t
qualities sequence line parses given qualities line in the context of a previously parsed sequence. The sequence is needed to assure the correct number of quality scores are provided. If not provided, this check is omitted.
val item_of_sexp : Sexplib.Sexp.t -> item
val sexp_of_item : item -> Sexplib.Sexp.t

Split a name string into a sequence identifier and an optional description. It is assumed that the given string is from an item's name field, i.e. that it doesn't contain a leading @ char.

Input/Output



Illumina-specific operations



1, 2, or 3

1 - 99, but usually 1 - 8

E.g. tile_of_string "2304" parses to

Inverse of tile_of_string.

Content of name lines as generated by Casava versions >= 1.8.

Parse a name string to a structured Illumina sequence_id. It is assumed that the given string is from an item's name field, i.e. that it doesn't contain a leading @ char.

Low-level Printing



This function converts item values to strings that can be dumped to a file, i.e. they contain full-lines, including all end-of-line characters.

Low-level Parsing



qualities sequence line parses given qualities line in the context of a previously parsed sequence. The sequence is needed to assure the correct number of quality scores are provided. If not provided, this check is omitted.
end