Clone wiki

ATLAS / Home

Welcome to ATLAS, your guide to the world of low-depth and ancient DNA!

ATLAS stands for Analysis Tools for Low-coverage and Ancient Samples. These tools cover all programs necessary to obtain variant calls, estimates of heterozygosity and more from a BAM file. There are sequence data processing tools, diagnostic tools, and variant discovery tools, similar to GATK by the Broad Institute.

For instructions on how to install and run ATLAS click here.

Engine parameters that are common to all tasks can be found here.

Each page of this wiki describes a different "task" implemented in ATLAS. Here is the full list of tasks implemented in ATLAS:

Sequence Data Processing Tools and Diagnostics

Variant Discovery Tools

Population Genetic Tools

Auxiliary Tools

Pipeline

We highly recommend using a local realignment tool on your BAM file and validating it with Picard-tools validateSamFile before taking off with ATLAS.

Pipeline for BAM files with single-end data:

  1. Split the read groups according to length with splitRGbyLength, use output file in subsequent steps
  2. Estimate the post-mortem damage patterns with estimatePMD
  3. If the genome is male, use X recal to recalibrate the base quality scores. If the genome is female, use BQSR or recal based on invariant sites to recalibrate.
  4. Use one of the variant discovery tools, one of the population genetics tools, or qualityTransformation. Take the base quality score recalibration and the PMD patterns into account by providing the corresponding files.

Pipeline for BAM files with paired-end data:

  1. Estimate the post-mortem damage patterns with estimatePMD
  2. If the genome is male, use X recal to recalibrate the base quality scores. If the genome is female, use BQSR or recal based on invariant sites to recalibrate.
  3. Create a BAM file with recalibrated quality scores with recalBAM, use this file in subsequent steps
  4. Merge read pairs using mergeReads, use output file in subsequent steps
  5. Re-estimate PMD patterns for the merged reads
  6. Use one of the variant discovery tools, one of the population genetics tools, or qualityTransformation. Take the second PMD patterns into account by providing the corresponding file.

In all tasks, reads that are not tagged as a "proper pair" by the SAM flag are considered to be single-end.

Questions?

See our Frequently Asked Questions or write an email to vivian.link@unifr.ch

Updated