Dear Martin, please when you have a minute, may I ask you for a suggestion : shall I have a VCF file as it is described below, please could you advise on a way to select only the variants with AD > 5, and AF > 0.05 ? thanks you very much ;)

#CHROM    POS    ID    REF    ALT    QUAL    FILTER    INFO    FORMAT    TUMOR    NORMAL
chr1    146139995    0    T    C    0    PASS    BaseCounts=1,5,0,38;ECNT=1;FS=4.639;GC=52.48;HCNT=2;HRun=0;LowMQ=0.0000,0.0682,44;MAX_ED=.;MIN_ED=.;NLOD=7.78;SOR=1.916;TLOD=7.51;VariantType=SNP;ANNOVAR_DATE=2016-02-01;Func.refGene=exonic;Gene.refGene=NBPF10;GeneDetail.refGene=.;ExonicFunc.refGene=nonsynonymous_SNV;AAChange.refGene=NBPF10:NM_001039703:exon5:c.A592G:p.K198E,NBPF10:NM_001302371:exon5:c.A592G:p.K198E;cosmic70=.;SIFT_score=.;SIFT_pred=.;Polyphen2_HDIV_score=.;Polyphen2_HDIV_pred=.;Polyphen2_HVAR_score=.;Polyphen2_HVAR_pred=.;LRT_score=.;LRT_pred=.;MutationTaster_score=.;MutationTaster_pred=.;MutationAssessor_score=.;MutationAssessor_pred=.;FATHMM_score=4.51;FATHMM_pred=T;PROVEAN_score=.;PROVEAN_pred=.;VEST3_score=0.26;CADD_raw=0.353;CADD_phred=6.192;DANN_score=0.130;fathmm-MKL_coding_score=0.000;fathmm-MKL_coding_pred=N;MetaSVM_score=-0.961;MetaSVM_pred=T;MetaLR_score=0.008;MetaLR_pred=T;integrated_fitCons_score=0.693;integrated_confidence_value=0;GERP�_RS=-0.47;phyloP7way_vertebrate=-1.190;phyloP20way_mammalian=-1.003;phastCons7way_vertebrate=0.002;phastCons20way_mammalian=0.004;SiPhy_29way_logOdds=3.32;dbscSNV_ADA_SCORE=.;dbscSNV_RF_SCORE=.;HRC_AF=.;HRC_AC=.;HRC_AN=.;HRC_non1000G_AF=.;HRC_non1000G_AC=.;HRC_non1000G_AN=.;esp6500siv2_ea=.;esp6500siv2_aa=.;esp6500siv2_all=.;ExAC_ALL=.;ExAC_AFR=.;ExAC_AMR=.;ExAC_EAS=.;ExAC_FIN=.;ExAC_NFE=.;ExAC_OTH=.;ExAC_SAS=.;ExAC_ALL=.;ExAC_AFR=.;ExAC_AMR=.;ExAC_EAS=.;ExAC_FIN=.;ExAC_NFE=.;ExAC_OTH=.;ExAC_SAS=.;ExAC_ALL=.;ExAC_AFR=.;ExAC_AMR=.;ExAC_EAS=.;ExAC_FIN=.;ExAC_NFE=.;ExAC_OTH=.;ExAC_SAS=.;Kaviar_AF=.;Kaviar_AC=.;Kaviar_AN=.;avsnp144=.;nci60=.;clinvar_20150629=.;ALLELE_END    GT:AD:AF:ALT_F1R2:ALT_F2R1:FOXOG:QSS:REF_F1R2:REF_F2R1    0/1:31,5:0.139:2:3:0.400:904,143:18:13    0/0:26,0:0.00:0:0:.:759,0:10:16

For a VCF object vcf, you'd like to select the rows with info(vcf)$AF > 0.05, which is a logical vector the same length as the number of rows in vcf.

AD is harder, but you'd like to select the appropriate column geno(vcf)$AD[,"TUMOR"]. This is a list(), one element for each row of the VCF, with the element equal to the AD of the corresponding genotype. You'd like to test whether the minimum value of each list element is greater than 5. An S4Vectors way of doing this is to coerce the plain-old list() to a List, and then exploit the fact that min() on the List returns element-wise minimums, min(as(geno(vcf)$AD[,"TUMOR"], "List")) >= 5. One might want to think carefully about this, e.g., if the variant allele is actually at high frequency.

A filter might be 

AFADfilter =function(x) {
    af <- info(x)$AF > 0.05
    ad <- min(as(geno(x)$AD[,"TUMOR"], "List")) >= 5
    af & ad
}

Create a FilterRules() instance (be sure to enclose the function in a list)

filter <- FilterRules(list(AFADfilter = AFADfilter))

To filter a VCF file, use the filter as the argument to filterVcf(),

vcf_filtered <- filterVcf(vcf_file, "hg38", "AF_and_AD_filtered.vcf", filters=filters)

To filter a VCF object, use

subsetByFilter(vcf, filter)