One　of　cancer　related　gene　function　research　is　the　discovery　of　gene　functional　module.　Here,　a　efficient　clustering　algorithm　proposed　to　reveal　the　gene　functional　modules　based　on　gene　expression　profiles.　First,　the　betweenness　algorithm　was　used　to　mine　the　cancer　gene　modules　from　the　158　feature　genes　extracted　from　the　20,　000　colon　cancer　gene　expression　profiles.　Second,　the　modularity　function　which　was　used　to　assess　the　performance　of　clustering　algorithm　was　applied　to　the　colon　cancer　gene　modules　and　got　7　gene　function　modules.　Finally,　combining　with　GO　database　the　biological　process　of　these　modules　was　classified.　The　results　show　that　this　algorithm　is　feasible　and　effective　in　gene　function　clustering.