skip to navigation
skip to content

Not Logged In

genmod 1.4

Annotate genetic inheritance models in variant files

Latest Version: 2.1.1

GENMOD
======

Tool for annotating patterns of inheritance Variant Call Format (VCF)
files.

Each variant in the VCF-file will be annotated with which genetic models
that are followed in the family. The genetic models that are checked are
the following:

- Autsomal Recessive, denoted 'AR'
- Autsomal Recessive denovo, denoted 'AR\_dn'
- Autsomal Dominant, 'AD'
- Autsomal Dominant denovo, 'AD\_dn'
- Autosomal Compound Heterozygote, 'AR\_comp'
- X-linked dominant, 'XD'
- X-linked dominant de novo, 'XD\_dn'
- X-linked Recessive, 'XR'
- X-linked Recessive de novo, 'XR\_dn'

**GENMOD** will add entrys to the INFO column for the given VCF file.
The new entrys are:

- **GM** A colon separated list with genetic models followed
- **ANN** Colon separated list with features overlapped in the
annotation file
- **Comp** Colon separated list with compound pairs(if any). These are
described like 'CHR\_POS\_REF\_ALT'.
- **MS** Model Score, a phred-score based on the genotype qualities to
describe the uncertainty of the genetic model.

All annotations will be present only if they have a value.

If ``-vep/--vep`` is used **ANN** will not be annotated since all
information is in the vep entry.

Installation:
-------------

**GENMOD** works with Python 2.7 and Python v3.2 and above

::

pip install genmod

or

::

git clone git@github.com:moonso/genmod.git
cd genmod
python setup.py install

USAGE:
------

Basic functions
~~~~~~~~~~~~~~~

::

genmod annotate ped_file variant_file

This will print a new vcf to standard out with all variants annotated
according to the statements above.

::

genmod build_annotation [--type] annotation_file

Genmod is distributed with a annotation database that are built from the
refGene data. If the user wants to build a new annotation set use the
command described above.

Alternatives
~~~~~~~~~~~~

- **GENMOD** can annotate the variants with 1000 genome frequencies.
Use the flag
``-kg/--thousand_g path/to/bgzipped/thousand_genomes.vcf.gz``
- Annotate with `CADD scores <http: cadd.gs.washington.edu=""/>`__, use
``-cadd/--cadd\_file path/to/huge_cadd_file.txt.gz``. There is also a
cadd file with all 1000 genomes positions (this one include some
indels), if annotation with this one use
``-c1kg/--cadd_1000_g path/to/CADD_1000g.txt.gz``
- If your VCF is already annotated with VEP, use ``-vep/--vep``
- If data is phased use ``-phased/--phased``
- If you want to allow compound pairs in intronic regions to use
``-gene/--whole_gene``
- If you want canonical splice site region to be bigger than 2 base
pairs on each side of the exons, use
``-splice/--splice_padding _integer_``
- The ``-strict/--strict`` flag tells **genmod** to only annotate
genetic models if they are proved by the data. If a variant is not
called in a family member it will not be annotated.

Distribution
~~~~~~~~~~~~

- GENMOD now includes db like files in the genmod/annotations folder,
this is the exon and gene definitions from
ftp://hgdownload.cse.ucsc.edu/goldenPath/hg19/database/refGene.txt.gz

If the user wants to use another annotation:

::

genmod build_annotation [--type] "annotation type" path/to/annotation_file -o path/to/outdir

In this case the new annotation will be built into the outdir specified
(default is the genmods annotation dir). When the user want to annotate
a vcf with this new annotation set use

::

genmod annotate ped_file variant_file [-a/--annotation_dir] /path/to/new/annotation_dir

- Compound heterozygote inheritance pattern will be checked if two
variants are exonic (or in canonical splice sites) and if they reside
in the same gene.

- GENMOD supports phased data use the -phased flag. Data should follow
the `GATK
way <http: gatkforums.broadinstitute.org="" discussion="" 45="" read-backed-phasing="">`__
of phasing.

- GENMOD support VCF files annotated with
`VEP <http: www.ensembl.org="" info="" docs="" tools="" vep="" index.html="">`__, use
-vep flag. This means that GENMOD will use the **VEP** annotation for
checking if variants are in the same gene.

- GENMOD can annotate variants with their
`CADD <http: cadd.gs.washington.edu=""/>`__ score. This is done by
adding the flag -cadd "path/to/cadd\_file".

Conditions for Genetic Models
-----------------------------

Short explanation of genotype calls in VCF format:
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Since we only look at humans, that are diploid, the genotypes represent
what we see on both alleles in a single position. 0 represents the
reference sequence, 1 is the first of the alternative alleles, 2 second
alternative and so on. If no phasing has been done the genotype is an
unordered pair on the form x/x, so 0/1 means that the individual is
heterozygote in this given position with the reference base on one of
the alleles and the first of the alternatives on the other. 2/2 means
that we see the second of the alternatives on both alleles. Some
chromosomes are only present in one copy in humans, here it is allowed
to only use a single digit to show the genotype. A 0 would mean
reference and 1 first of alternatives.

If phasing has been done the pairs are not unordered anymore and the
delimiter is then changed to '\|', so one can be heterozygote in two
ways; 0\|1 or 1\|0.

Autosomal Recessive
~~~~~~~~~~~~~~~~~~~

For this model individuals can be carriers so healthy individuals can be
heterozygous. Both alleles need to have the variant for an individual to
be sick so a healthy individual can not be homozygous alternative and a
sick individual *has* to be homozygous alternative.

- Affected individuals have to be homozygous alternative (hom. alt.)
- Healthy individuals cannot be hom. alt.
- Variant is considered *de novo* if both parents are genotyped and do
not carry the variant

Autosomal Dominant
~~~~~~~~~~~~~~~~~~

- Affected individuals have to be heterozygous (het.)
- Healthy individuals cannot have the alternative variant
- Variant is considered *de novo* if both parents are genotyped and do
not carry the variant

Autosomal Compound Heterozygote
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

This model includes pairs of exonic variants that are present within the
same gene. **The default behaviour of GENMOD is to look for compounds
only in exonic/canonical splice sites**. The reason for this is that
since some genes have huge intronic regions the data will be drowned in
compound pairs. If the user wants all variants in genes checked use the
flag -gene/--whole\_gene.

1. Non-phased data:

- Affected individuals have to be het. for both variants
- Healthy individuals can be het. for one of the variants but cannot
have both variants
- Variant is considered *de novo* if only one or no variant is found
in the parents

2. Phased data:

- All affected individuals have to be het. for both variants **and**
the variants has to be on two different alleles
- Healthy individuals can be heterozygous for one but cannot have
both variants
- If only one or no variant is found in parents it is considered *de
novo*

X-Linked Dominant
~~~~~~~~~~~~~~~~~

These traits are inherited on the x-chromosome, of which men have one
allele and women have two.

- Variant has to be on chromosome X
- Affected individuals have to be het. or hom. alt.
- Healthy males cannot carry the variant
- Healthy females can carry the variant (because of X inactivation)
- If sex is male the variant is considered *de novo* if mother is
genotyped and does not carry the variant
- If sex is female variant is considered *de novo* if none of the
parents carry the variant

X Linked Recessive
~~~~~~~~~~~~~~~~~~

- Variant has to be on chromosome X
- Affected males have to be het. or hom. alt. (het is theoretically not
possible in males, but can occur due to Pseudo Autosomal Regions).
- Affected females have to be hom. alt.
- Healthy females cannot be hom. alt.
- Healthy males cannot carry the variant
- If sex is male the variant is considered *de novo* if mother is
genotyped and does not carry the variant
- If sex is female variant is considered *de novo* if not both parents
carry the variant

Details
-------

Exonic variants
~~~~~~~~~~~~~~~

Variants are defined as exonic if they are within an interval that is
defined as an exon in the annotation file, and if they are in the
canonical splice sites. The size of the canonical splice sites can be
altered with -splice 'integer', in this case the annotation needs to be
rebuilded. Example:

::

run_genmod ped_file variant_file -an refGene.txt -at gene_pred -splice 6

In this case the exons will be padded by 6 bases on each side.

If VEP annotation is used the following SO-terms is counted as compound
candidate sites:

**transcript ablation, splice donor variant, splice acceptor variant,
stop gained, frameshift variant, stop lost, initiator codon variant,
inframe insertion, inframe deletion, missense variant, transcript
amplification, splice region variant,incomplete terminal codon variant,
synonymous variant, stop retained variant, coding sequence variant**.

Annotation Formats
~~~~~~~~~~~~~~~~~~

The following formats are supported:

- `gene pred <http: genome.ucsc.edu="" faq="" faqformat.html#format9="">`__.
This is the format of the refSeq genes
- `gtf <http: www.ensembl.org="" info="" website="" upload="" gff.html="">`__. This
is the format use by ensembl
- CCDS format
- `BED <http: genome.ucsc.edu="" faq="" faqformat.html#format1="">`__\ format.

When BED format is used all entrys will both count as exons and
genes(for compounds).

.. raw:: html




 
File Type Py Version Uploaded on Size
genmod-1.4.tar.gz (md5) Source 2014-07-04 8MB
  • Downloads (All Versions):
  • 28 downloads in the last day
  • 1102 downloads in the last week
  • 8597 downloads in the last month