Since you tagged it as R I'll add an R answer.

Example data.

df <- data.frame(HGVS.Consequence=c("Met1?", "Phe12Ser", "Ala2Glu"))
> df
  HGVS.Consequence
1            Met1?
2         Phe12Ser
3            Ala2Glu

Tidyverse answer.

library("tidyr")

extract(
  df, HGVS.Consequence, into=c("aa1", "pos", "aa2"),
  regex="(^[A-Z][a-z]+|\\?)([[:digit:]]+)([A-Z][a-z]+|\\?)")

  aa1 pos aa2
1 Met   1   ?
2 Phe  12 Ser
3 Ala   2 Glu

sed 's/^\([^0-9]*\)\([0-9]*\)\([^0-9]*\)$/\1\t\2\t\3/' < in > out

Here is Perl one-liner solution :

echo "Met1?"|perl -ne 's/(\D+)(\d+)(\D+)/\1 \2 \3/g;  print'

Met 1 ?

 $ echo "Ser25Phe" | sed -r 's/^([^[:digit:]]+)([[:digit:]]+)([^[:digit:]]+)$/\1\t\2\t\3/'

Ser 25  Phe

 $ echo "Ser25Phe"| while read line; do echo ${line%%[0-9]*}"\t"${line//[^0-9]}"\t"${line##*[0-9]};done

Ser 25  Phe

with R:

> df <- data.frame(HGVS.Consequence=c("Met1?", "Phe12Ser", "Ala2Glu","?1Met"))
> library(stringr)
> library(magrittr)
> str_extract_all(df$HGVS_Consequence,"\\D+|\\d+", simplify = T) %>%
+     set_colnames(c("Before", "Position", "After")) 

     Before Position After
[1,] "Met"  "1"      "?"  
[2,] "Phe"  "12"     "Ser"
[3,] "Ala"  "2"      "Glu"
[4,] "?"    "1"      "Met"

AsAsp line/record will not be parsed this way.