Inactivation of a specific gen can occur by a site specific methylation of CpGs in the promoter area of the gene or more simply by a vector producing RNAi. Alternatively one can use genome editing to cut out and eliminate the gene.
I'd suggest you start with a literature review. FYI, there are a LOT of terms for RNA-based gene silencing (RNAi, quelling, knock-down, etc) and many of them can activate methylation-based gene silencing.
Are you trying to get a reduced-transcription variant of a particular gene in a particular organism? If so, you should make use of the existing mutant lines available (from published papers, other labs, stock centers, etc.). No need to duplicate effort, especially not for a genetic screen.
Other methods: T-DNA transformation often but not always gives a knockout or knockdown phenotype (order mutants from the stock center); random mutagenesis (again, order the mutant from a stock center); allele variant in the population (e.g. an ecotype) that happens to have lower expression.
Targeting RNA or protein transiently (using non-integrating transient RNAi tools or plasmids or targeted protein-degradation) could give short-term (days) incomplete depletion. You could also generate conditionally-driven stable cell-lines to have long-term silencing when required.
Targeting functional DNA elements (Promoters/CpG islands/TSS/Exons with CRISPR) may have a permanent effect on the gene activity.