Background

The contribution of pathogen genetic variation to the risk of developing lower

respiratory tract infection(LRTI) remains incompletely understood. We used genome

wide association study(GWAS) cohort to examine the contribution of genetic variations

to the risk of LRTI amongst African children enrolled in an intensively sampled birth

cohort.

Methods

Nasopharyngeal(NP) swabs were collected 2-weekly from 800 infants over the first year

of life. We analysed 393 and 777 pneumococcal genomes sampled from infants with

and without LRTI. We performed univariate linear mixed model GWAS(FaST-LMM) after

filtering out variants with missingness >5% and minor allele frequency(MAF) <1%.

Results

We identified 100,504 and 12,316,614 total SNPs and k-mers. Of these, 48,607 SNPs

and 389,918 unique k-mer patterns remained after quality control. GWAS analysis of

the pre-processed variants revealed statistically significant associationqs(- value<0.05)

in 17 SNPs and 3,237 k-mers (tagged by 64 unique k-mers) after Bonferroni correction.

Overall, 94%(16/17) SNPs and 84%(2,733/3,273 k-mers) were located in genic regions.

The variant-containing genes were diverse including surface-exposed,

transcription/response regulators and sugar/nutrient transporters.

Conclusions

Our preliminary analysis suggests that pathogen genetic variation may contribute to the

risk for LRTI but further analyses are required to elucidate the biological mechanisms

and degree of this contribution.