Let$\DeclareMathOperator\MCG{MCG}\DeclareMathOperator\vcd{vcd}\DeclareMathOperator\cd{cd}$Let $B_{k}(S_{g}),$ $MCG(S_{g};k)$$\MCG(S_{g};k)$ and $MCG(S_{g}))$ are Braid$\MCG(S_{g}))$ denote the braid group, Mappingthe mapping class group (relative to $k$) and Mappingthe mapping class group of the orientable surface $S_{g}$, respectively. For $g\geq3,$ we have a short exact sequence $$1\longrightarrow B_{k}(S_{g})\longrightarrow MCG(S_{g};k)\longrightarrow MCG(S_{g})\longrightarrow 1.$$$$1\longrightarrow B_{k}(S_{g})\longrightarrow \MCG(S_{g};k)\longrightarrow \MCG(S_{g})\longrightarrow 1.$$ This short exact sequence gives the relationalrelation between the (virtual) cohomological dimensions: $$vcd(MCG(S_{g};k))\leq cd(B_{k}(S_{g}))+vcd(MCG(S_{g})).$$ Is$$\vcd(\MCG(S_{g};k))\leq \cd(B_{k}(S_{g}))+\vcd(\MCG(S_{g})).$$ Does the same type of relation exist for punctured orientable surface $S_{g}-\{p_{1},\ldots,p_{n}\}?$