HOGImine

HOGImine (Higher-Order Genetic Interaction miner), is a pattern-mining-based algorithm for finding genetic meta-markers, i.e. combinations of genetic markers, that show a statistical association with a phenotype. Compared to the state-of-the-art miners, it expands the class of discoverable genetic meta-markers by considering higher-order interactions of genes and by allowing multiple encodings for the genetic variants. Moreover, our method can exploit prior biological knowledge on gene interactions, such as protein-protein interaction networks, genetic pathways and protein complexes, to restrict its search space. HOGImine accepts both binary encodings and additive encodings for the markers.

Examples of meta-markers considered by HOGImine. (a) The dataset features, for each sample, a vector of genetic markers with additive encoding. The markers are subdivided into genes, in the figure highlighted by pale colors. (b) Information on the interactions among genes if usually provided via a PPI network. (c) The black solid meta-marker (shown in Panel a) spans three genomic intervals in the violet, yellow and azure genes, which form a connected subgraph in the network, using the dominant encoding for all three. (d) The black dashed meta-marker spans three genomic intervals in the yellow, green and red genes, which form a connected subgraph in the network, using the recessive encoding for all three.

Citing our work

HOGImine is described in the following paper:

Paolo Pellizzoni, Giulia Muzio and Karsten Borgwardt. Higher-order genetic interaction discovery with network-based biological priors. ISMB 2023. [PDF online]

Compilation

Compilation uses make. Currently the code is compiled with gcc-12, but other compilers might work as well.

cd src
make

To decompress the SNP data in the data/mmusculus folder, use the following command.

unzip data/mmusculus/BMC.Mode/BMC.Mode_X_T.txt.zip -d data/mmusculus/BMC.Mode/

Options

The options for HOGImine are:

• -i file: marker (binary or additive encoding) file. It should be matrix of size $n. SNPs \times n. samples$, i.e. with each column being a sample.
• -l file: labels file. It should be a binary matrix of size $n. samples \times 1$.
• -s file: SNP names file. It should be a file with $n. SNPs$ rows, containing the names of the SNPs.
• -c file: covariate file. It should be a matrix of size $n. samples \times 1$ taking values in ${ 1, \dots, C }$, with $C$ being the number of covariate classes.
• -m file: SNP map file. It should be a file with $n. genes$ rows. The first entry of each row should be the the gene name, and the following entries should be the SNPs that are mapped to that gene.
• -e file: gene interactions file. It should be a file with $n. interactions$ rows. Each row should contain a set of genes that are known to be interacting (biological priors).
• -f level: target fwer (e.g. 0.05).
• -o or -O file: output file name, with -o in compressed format and with -O in verbose format.
• -p p: number $p$ of permutations (default $0$), if $p > 0$ it runs a permutation testing procedure.
• -d d: maximum SNP interval length in each gene.
• -v: outputs all the testable patterns.

Usage example (binary data)

./src/hogimine_additive -i data/athaliana/interactome_0kb/avrRpm1/avrRpm1_X.txt -l data/athaliana/interactome_0kb/avrRpm1/avrRpm1_Y.txt -c data/athaliana/covar_snps/avrRpm1/avrRpm1_covar_n2.txt -s data/athaliana/interactome_0kb/avrRpm1/avrRpm1_snpID.txt -m data/athaliana/interactome_0kb/avrRpm1/avrRpm1_snp_map.txt -e data/athaliana/athal_ppi/genes+edges+3cliques+complexes.txt -f 0.05 -O out_athaliana

Output: each pattern is reported on a row, the first column reports the CMH p-value of the pattern, then the interacting genes spanned by the pattern are reported, and finally the SNP positions composing the pattern.

3.27815e-12; AT3G07040, AT3G25070; Chr3_2227817, Chr3_9133148
1.33438e-11; AT3G07040; Chr3_2227817
2.22304e-11; AT3G07040, AT3G25070; Chr3_2227817, Chr3_9132644
2.22304e-11; AT3G07040, AT5G51450; Chr3_2227817, Chr5_20897395
...

Usage example (additive data)

./src/hogimine_additive -i data/mmusculus/BMC.Mode/BMC.Mode_X_T.txt -l data/mmusculus/BMC.Mode/BMC.Mode_y_bin.txt -c data/mmusculus/BMC.Mode/BMC.Mode_kinship.txt -s data/mmusculus/BMC.Mode/SNPs_ids.txt -m data/mmusculus/BMC.Mode/gene2snps.txt -e data/mmusculus/BMC.Mode/all_int_compl.txt -f 0.05 -O out_mus

Output: each pattern is reported on a row, the first column reports the CMH p-value of the pattern, then the interacting genes spanned by the pattern are reported along with the encoding for the SNPs, and finally the SNP positions composing the pattern.

3.65197e-25; ENSMUSG00000047040 (dominant), ENSMUSG00000048175 (recessive); chr11_96928685, chr15_98155192
3.65197e-25; ENSMUSG00000047040 (dominant), ENSMUSG00000048175 (recessive); chr11_96930807, chr15_98157791
...
4.58913e-25; ENSMUSG00000061298 (dominant), ENSMUSG00000038756 (dominant); chr4_110873242, chr11_96148817
...
1.76405e-24; ENSMUSG00000057058 (dominant), ENSMUSG00000022148 (recessive); chr11_96573406, chr11_96584561, chr15_6571463
1.8372e-24; ENSMUSG00000018672 (dominant), ENSMUSG00000060992 (recessive); chr11_96859685, chr15_103292285
...

Acknowledgements

This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement No 813533 (K.B.).