Summing Based on Part of a Control Cell

I assume that Eszter would like a list of totals for all the mutation codes, rather than recording them one at a time. That is, to use an array of unique codes as the second argument in Sumif.
For reasons of simplicity, format the data as a table. Change Column1 heading to "Codes" and Column2 heading to "Number". Insert one table column to the right, change its heading to "Prefix"
In the first row under "Prefix" enter =TEXTBEFORE([@Code]]," "). You don't have to type [@Code], just click on the first value.

In a nearby cell enter =UNIQUE(Table1[Prefix]). In my test it was at E2.
In a cell in an adjacent column enter =SUMIF(Table1[Prefix],E2#,Table1[Number]), Where E2 is the address of the first unique prefix in my test.

Job done.
(see Figure 1 below)

Here's an interesting problem. What if the mutation codes did not have a space to separate each preface code from the remaining digits? Suppose the mutation codes were like AKT142 or BRAF1975. How would you determine the preface codes?
The following function in My Excel Toolbox returns an array of numeric values beginning with Start incrementing by Step until Finish (not beyond):
    =ForNext(Start, Finish, [Step], [AsColumn])
Start, Finish, and Step can be any numeric value. Default Step is 1.
If AsColumn is FALSE (default), the result is a 1D row array; TRUE returns a 2D column array. Expect N elements: N = 1 + INT((Finish - Start) / Step)
Using ForNext, this formula returns the preface code for cell A1:
    =LEFT(A1, SUM(--ISERROR(VALUE(MID(A1, ForNext(1, LEN(A1)), 1)))))
MID(...) returns an array containing each character. VALUE(...) returns an error if the character is not a digit (0,1,2...). Double unary (--) converts TRUE/FALSE into 1/0. So the formula returns AKT if cell A1 is AKT142 or BRAF if it is BRAF1975.
This formula uses the SEQUENCE function in Excel 2021 to return the same result:
    =LEFT(A1, SUM(--ISERROR(VALUE(MID(A1, SEQUENCE(LEN(A1)), 1)))))
For older versions of Excel,
    SEQUENCE(LEN(A1))
can be replaced by
    ROW(INDIRECT("1:" & LEN(A1)))
My Excel Toolbox also includes the following regular expression function:
    =RegExMatch(Text, Pattern, [Mode], [IgnoreCase], [Multiline])
This is similar to Excel's new REGEXEXTRACT function described here: https://support.microsoft.com/Search/results?query=REGEXEXTRACT
Therefore, these formulas also return the preface code for cell A1 (which is assumed to be upper case):
    =RegExMatch(A1, "[A-Z]+")
    =REGEXEXTRACT(A1, "[A-Z]+")
These functions use the VBScript 5.5 regular expression syntax described here:
https://learn.microsoft.com/en-us/previous-versions/1400241x(v=vs.85)
Or here: https://drive.google.com/file/d/1qWS2qLib0sL_u6Sh95FuEeZxRzzr4SUE/view
See https://sites.google.com/view/MyExcelToolbox/

The Tip assumes the mutation codes in column A do not begin with a space and each preface code (like AKT or BRAF) ends with a space. It says, "There are many ways you could approach this problem...."
Re. Using a Helper Column:
If column C is the helper column and the sum is desired for a preface code in cell $E$1 (which must not include a space), then use this formula:
    =SUMIF(C:C, $E$1, B:B)
Re. Using SUMPRODUCT:
This formula works just as well as the SUMPRODUCT formula:
    =SUM(--(LEFT(A:A, LEN($E$1)) = $E$1) * B:B)
Again, $E$1 must not include a space.
Re. Using SUMIF Directly:
Since the preface code ends with a space, the Tip's formulas containing an asterisk should take this into account. Therefore, they should have a space before the asterisk like this:
    =SUMIF(A:A, "AKT *", B:B)
    =SUMIF(A:A, $E$1 & " *", B:B)
Otherwise, there would be no difference between preface codes AKT and AKTX, for example, or between BRAF and BRAFX. As before, $E$1 must not include a space.