|
Alta-Cyclic
A self optimized base caller for Illumina/Solexa Genome Analyzer (patent pending)
|
Designing the experiment:
Make sure that you have on your flow-cell at least one rich DNA library with a known reference genome.
We usually use phi-X174. You can download its reference genome in 2bit format here.
After the run:
Analyze the image files with Firecrest and call the reference library with Bustard.
Launch Atla-Cyclic Training:
Type brave_heart.pl [MANDATORY PARAMETERS] [OPTIONAL PARAMETERS]
Launch Atla-Cyclic Calling:
Type karma.pl --auto [OPTIONAL PARAMETERS]
brave_heart - A trainer program for Alta-Cyclic base caller.
SYNOPSIS
brave_heart.pl [MANDATORY PARAMETERS] [OPTIONAL PARAMETERS]
brave_heart.pl --load Path_to_Training_Folder_of_Broken_Run/egg
OPTIONS
Mandatory Parameters:
--number_of_cycles Number of cycles in the run
--path_to_intensity_files Path to the folder that contains the intensity files for training
--path_to_training_sequences Path to the folder that contains sequences that corresponds to intensity files
--lane_for_training Lane number that is used for the training
--tiles_for_training The range of tiles that will be used for training
--reference_genome Full path to the reference genome
--is_reference_genome_local Y - indicates that the refernce genome should be copied to the cluster nodes
N - assumes that the genome is installed on the cluster nodes
Optional Parameters:
--trim Trim the trainings sequences for a certain number. Makes the training faster
--threshold_for_analysis Set the threshold for allowing mismatches in the training seqeunces
--maximum_lines_for_training Set the intial amount of eneries that will be sampled for training
--flowcell The flowcell name
--verbose Set verbosity
--zipped If the intensity files are zipped
--debug Give extra messages and indicators. Stores .o files that return from the cluster
--not_nice Do not report your optimization results to Alta-Cyclic developers. This may prevent us of collecting statistics and improve later versions.
Miscellaneous:
--load Take an 'egg' file from a broken training run, and continue the run from the placed it stopped.
Help Options:
--help Show a list of parameters.
--man Print a long manual.
EXAMPLE
brave_heart.pl --number_of_cycles 78 --path_to_intensity_files ~/int_from_the_classic_run --path_to_training_sequences ~/sequences_for_training_Classic/ --lane_for_training 8 --tiles_for_training 1-100 --reference_genome ~/DBase/phiX.2bit --is_reference_genome_local Y --trim 25 --threshold_for_analysis 20 --maximum_lines_for_training 100000 --verbose --debug --flowcell
FLOWCELL_1234 --trim 25
karma - The base caller part of Alta-Cyclic
SYNOPSIS
karma.pl --auto [--tiles --lanes]karma.pl [MANDATORY PARAMETERS] [OPTIONAL PARAMETERS]
OPTIONS
| --auto | Gets the parameters from brave_heart. Only under Karma folder. NOTICE: the default tile range is 1-100. On a GAI machine you should use the --tiles option. See example below |
| Mandatory Parameters: | |
| --random_walk | Expect to get [best p1] and [best p2] that were found by brave_heart |
| --number_of_cycles | Number of cycles in the run |
| --path_to_intensity_files | Path to the folder that contains the intensity files for reading |
| --number_of_cycles | Path to the folder that contains the intensity files for training |
| --lanes | Lanes to read |
| --tiles | Tiles to read |
| --SVM_tar_location | Location of the SVM package that was generated by brave_heart |
| Optional Parameters: | |
| --zipped | If the intensity files are zipped |
| --flowcell | The flowcell name. |
| --verbose | Set verbosity. |
| --debug | Give extra messages and indicators. Stores .o files that return from the cluster |
| Help Options: | |
| --help | Show a list of parameters. |
| --man | Print a long manual. |
EXAMPLES
./karma.pl --auto |
Your results are under the Karma directory in the same format that Illumina reports.