Examples

In these examples, we run SOPRANO with SSB7. We will use the readily available example annotated somatic mutation and bed protein input files shipped with SOPRANO. These inputs are compatible with the GRCh37 reference genome. We will seed the value of the simulations such that the outputs described on this page can be reproduced on your own computer.

The main entry point for executing the soprano pipeline is soprano-run.

Note: If you have installed SOPRANO with the mpi or dev flavours, you can prefix this with your appropriate MPI executable and processor count. For example, mpirun -n X, mpiexec -n X, srun -n X --pmix, and so on; for distribution over X processors.

---

Example 1: dN/dS with analytic approximations to uncertainty

The most straightforward analysis involves running SOPRANO against a single file, to compute dN/dS and estimate the statistical uncertainty via Katz method.

In this case, one can execute soprano-run @example_01.soprano

example_01.soprano

--input=/path/to/SOPRANO/data/example_annotations/TCGA-05-4396-01A-21D-1855-08.annotated
--bed_file=/path/to/SOPRANO/data/example_immunopeptidomes/TCGA-05-4396.Expressed.IEDBpeps.SB.epitope.bed
--name=TCGA-05-4396-example_01
--output=/tmp
--assembly=GRCh37

This will cache the results into the output directory /tmp/TCGA-05-4396-example_01. There should only be a single sub-folder, data. The results from this analysis are then inside the file /tmp/TCGA-05-4396-example_01/data/data.results.tsv.

Coverage	ON_dnds	ON_Low_CI	ON_High_CI	ON_Mutations	OFF_dNdS	OFF_Low_CI	OFF_High_CI	OFF_Mutations	Pvalue	ON_na	ON_NA	ON_ns	ON_NS	OFF_na	OFF_NA	OFF_ns	OFF_NS
Exonic_Only	0.17320730117340286	0.031724321858695226	0.9456709370621603	6.0	0.866868735958057	0.49699058713838856	1.5120234161945458	63.0	0.34485641679719375	2.0	1027780.0	4.0	356038.0	46.0	10627300.0	17.0	3404610.0
Exonic_Intronic	0.17320730117340286	0.031724321858695226	0.9456709370621603	6.0	0.866868735958057	0.49699058713838856	1.5120234161945458	63.0	0.34485641679719375	2.0	1027780.0	4.0	356038.0	46.0	10627300.0	17.0	3404610.0

Example 2: dN/dS with numerical estimates of uncertainty

Alternative statistical measures on the significand of dN/dS can be obtained via generating a suite of null hypothesis simulations from randomization of the bed protein genomic background. In this case, we should specify some number of samples to the SOPRANO definition file

In this case, one can execute soprano-run @example_02.soprano

example_01.soprano

--input=/path/to/SOPRANO/data/example_annotations/TCGA-05-4396-01A-21D-1855-08.annotated
--bed_file=/path/to/SOPRANO/data/example_immunopeptidomes/TCGA-05-4396.Expressed.IEDBpeps.SB.epitope.bed
--name=TCGA-05-4396-example_02
--output=/tmp
--assembly=GRCh37
--seed=333
--n_samples=19

This will cache the results into the output directory /tmp/TCGA-05-4396-example_02. There should be the same content as generated in the previous example, in addition to a statistical summary file from in addition to an automatically generated an accompanying figure.pdf file.

statistics.json

# Placeholder
{
    "ON_dNdS": {
        "exonic_only": {
            "samples": {
                "mean_value": 0.7721069499148594,
                "std_dev": 0.04871517355356516
            },
            "data": {
                "value": 0.7028932965124453,
                "p_value_left": 0.08982714832397816,
                "p_value_right": 0.9101728516236403
            }
        }
    },
    "OFF_dNdS": {
        "exonic_only": {
            "samples": {
                "mean_value": 0.8301892065970186,
                "std_dev": 0.004196696763633817
            },
            "data": {
                "value": 0.837001816221436,
                "p_value_left": 0.9487837348903762,
                "p_value_right": 0.05121626510654794
            }
        }
    }
}